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!

STATISTICAL ANALYSIS about the wild MVP RACE and HARDEN/CURRY vs PREVIOUS MVP WINNERS (VOA)

Discussion in 'Houston Rockets: Game Action & Roster Moves' started by Voice of Aus, Mar 25, 2015.

  1. Voice of Aus

    Voice of Aus Contributing Member

    Joined:
    Jun 28, 2013
    Messages:
    5,150
    Likes Received:
    402
    Introduction

    Its been a while since I did a stats based thread on clutchfans, with my last one being in the offseason discussing the value of Ariza's vs Parson's respective contratcs and overall performance along with other SF’s http://bbs.clutchfans.net/showthread.php?t=256961. However now with all eyes on the postseason now and now a large enough sample size to separate players I thought it would be a good time create my MVP race stats breakdown.

    2014-2015 Season Candidates

    At first I wanted to start out with this years fab 5 candidates discussed in the media which are Harden, Curry, James, Westbrock and Davis however with further research and then hearing comments from respected NBA brains such as Zach Lowe decided to introduce a 6th member into the MVP race in Chris Paul.

    My Main objective was to try discover ways to utilise both player performance and team success/failure in regard to if the player has played in those games.

    To find out player performance I took into 3 major advanced stats being
    - Win Shares
    - RPM
    - VOR (value over replacement player)

    And the equation that was basic to create what im calling Total Player Peformance (TPP)
    TPP=(WS*RPM*VOR )/100

    To find out team performance in relation to if the player has played I used Overall Success (OS)
    OS= team win% * (Games played/total team games played)

    thus generating these results for the 6 candidates

    [​IMG]

    which then formed the framework of the graphical findings

    [​IMG]

    As you can see it looks like a 2 horse race in the race for the MVP for Curry and Harden who both have significant claims statistically and on the eye test, Curry's debate has to centre more around team success as his overall play with using TPP is slightly below that of Harden who is having a career year, however as you will find out later has a historically low OS for an MVP to have.

    Then as I expected Chris Paul is right their to make a strong claim for 3rd/4th despite no one in the media talking about his year.

    Then James whose year is either the end of domination lebron or he’s coasting for now and he’s going to have so much left in the tank for the playoffs compared to his previous years because this year has been somewhat disappointing for him statistically.

    Lastly we have russ and davis who seem to have a mixture of missed games and their teams just haven’t won enough.

    How Curry and Harden rank amongst previous MVP Winners.

    I was really more interested in finding this out and seeing if their even is a worthy winner in the field this year. Using the same data (TPP and OS) you can see the spread in winners.

    [​IMG]

    and then also graphically

    [​IMG]

    First off, Lebron was really really good, scary good when you think about it, his 08-09 season is one of the best seasons ever by a player in the NBA.

    But when you see this graph you start to see why Curry is a choice the media seems to like to make, when you take into account that the Warriors are going to finish with around 68 wins his season and MVP comparison should be a Dirk 06-7 esq year with a tremendous winning record and attendance in games but just a good-great year individually. (basically nest player on a great team).

    It’s a shame we couldn’t of won a few more games because if harden keeps up his current 2014-15 TPP his season is going to almost mirror that of Durant last year and we all know how easily he won the award in the end, it seemed like the sexy choice by all in the media having lebron MVP fatigue.

    Conclusion

    Harden, Curry, Harden, Curry even with all these stats thiers still not a obvious choice, going with recent trends its gotta be curry as the media values wins a lot more than myself but that’s here nor there. When breaking it down to its simplist form I generally don’t know yet and hope the remaining 20-22 remiming games can make some clarity such as curry resting for 5-10 games would help a lot ;) or the beard just carrying us with his friend D12 now back in the fold should be able to create more wins for us as we battle for home court.

    ill try and answer any questions, concerns or disagreements if their happens to be any, thanks VOA

    [​IMG]


    TL;TD: this isn’t the thread you’ll find an answer in but its down to curry and harden and the remaining 22 games count for more than ever in an MVP race.
     
    #1 Voice of Aus, Mar 25, 2015
    Last edited: Mar 25, 2015
  2. JCDenton

    JCDenton Member

    Joined:
    Oct 10, 2007
    Messages:
    1,033
    Likes Received:
    163
    This statistical analysis has not been PEER REVIEWED by me or another statistical genius prior to your illicit publication. Please delete this post immediately or I will be forced to file a lawsuit against you for the crime of STATISTICAL ABUSE.
     
    1 person likes this.
  3. FTW Rockets FTW

    FTW Rockets FTW Contributing Member
    Supporting Member

    Joined:
    Jun 23, 2011
    Messages:
    27,724
    Likes Received:
    21,360
    I am Milton Friedman's grandson and I was born with ingrained, complex statistics. I APPROVE this thread.
     
    1 person likes this.
  4. Rocket2008

    Rocket2008 Member

    Joined:
    Feb 18, 2015
    Messages:
    1,441
    Likes Received:
    37
    It's going to be Curry, I'd be really surprised if it was Harden. Also, I cannot really be mad at that, because let's face it: if the Rockets would be beating teams the way the Warriors are with a franchise player who is doing great in a lot of categories and on pace to win over 65 games, how can you not reward that? Harden is already recognised by plenty of players and media so I'm actually okay with Curry winning. I rather Harden win it in a year where we have a better system, Morey and Leslie decide to go all-in and we can be confident about going to at least the WCF.
     
  5. Voice of Aus

    Voice of Aus Contributing Member

    Joined:
    Jun 28, 2013
    Messages:
    5,150
    Likes Received:
    402
    how weird I'm John Maynard Kaynes great grandson (never thought i would ever see economics jokes in the GARM) :grin:
     
  6. chenjy9

    chenjy9 Numbers Don't Lie
    Supporting Member

    Joined:
    Nov 1, 2008
    Messages:
    9,097
    Likes Received:
    3,816
    /argument

    #Harden4MVP
     
  7. Voice of Aus

    Voice of Aus Contributing Member

    Joined:
    Jun 28, 2013
    Messages:
    5,150
    Likes Received:
    402
    taking away success and games played the race is still extremely tight up the top.
    i have harden with a TPP of 8.95 and Curry with a 8.02 which would be higher had he not missed 15 4th quarters completely due to blow out wins.

    but i don't mind that hashtag tho
     
  8. blahblehblah

    blahblehblah Member

    Joined:
    Mar 29, 2006
    Messages:
    2,040
    Likes Received:
    206
    Nice work, really enjoyed the graphs.

    One suggestion. Instead of using TWO cumulative metric (VORP, WAR) and ONE per minute metric (RPM), it might be better to add something like Win Shares per 48 or WAR per 48 to balance the data.

    This way a player who plays every game and 36+ minutes is rewarded, while at the same time the per minute efficiency/production of Westbrook, ADavis & Curry are also equally rewarded.

    08-09 Lebron !!
     
  9. Voice of Aus

    Voice of Aus Contributing Member

    Joined:
    Jun 28, 2013
    Messages:
    5,150
    Likes Received:
    402
    thanks for the feedback, i avoided the per 48 stats because i wanted to give significant value to players playing games and not missing games like russ/ davis and LBJ compared to curry/ harden/ CP3 who have all played over 90% of available games, but i see your point, ill run the numbers later with per.48's and see how it comes out
     
  10. blahblehblah

    blahblehblah Member

    Joined:
    Mar 29, 2006
    Messages:
    2,040
    Likes Received:
    206
    Also not sure if multiplying the differing metrics is an accurate way to determine value. Each metric value is distinct from the other, meaning their score isn't weighed the same, and by multiplying them, metrics with higher scores are inflated while those with lower numerical scores are deflated.
     
  11. heypartner

    heypartner Contributing Member

    Joined:
    Oct 27, 1999
    Messages:
    53,634
    Likes Received:
    30,829
    First, shouldn't WS * VORP * RPM be well into the hundreds.

    Regardless, I don't see how you can just wave a magic wand and decide that multiplying those three stats equals a new stat. I think we're going to need some kind of made up coefficient to put into that formula. :D

    You're actually multiplying the same number more than once across those three formulas...thus you are skewing the importance of one variable vs others outside of any statistical modelling that the inventors intended.

    Seriously though, why not simply overlay the three ontop of each other and we can see who performs the best per established stat. Why invent a new one?
     
  12. AirBud#10

    AirBud#10 Member

    Joined:
    Sep 14, 2010
    Messages:
    2,729
    Likes Received:
    202
    The only way Harden wins is if the Rockets win out and the Warriors finish with 65 or fewer wins. I don't really see that happening unfortunately.
     
  13. Voice of Aus

    Voice of Aus Contributing Member

    Joined:
    Jun 28, 2013
    Messages:
    5,150
    Likes Received:
    402
    I was bored on a flight home and wanted to know who the MVP should be, I did miss off tho one bit of information you highlighted, I divided everything by 100 to make easier numbers to demonstrate on a graph for the TPP.
     
  14. RudyT43

    RudyT43 Member

    Joined:
    Jun 28, 2014
    Messages:
    423
    Likes Received:
    40
    Liked the post. Helps us clutchfans have more of a reason to hate ESPN.
     
  15. RedDragon01

    RedDragon01 Member

    Joined:
    Oct 13, 2009
    Messages:
    560
    Likes Received:
    81
    This was my first thought while reading over it.
     
  16. cheke64

    cheke64 Member

    Joined:
    May 12, 2009
    Messages:
    16,070
    Likes Received:
    3,674
    The drunken MVP master. Hidden defense crouching beard
     
  17. JayZ750

    JayZ750 Contributing Member

    Joined:
    May 16, 2000
    Messages:
    20,014
    Likes Received:
    3,599
    You randomly multiplied three stats.
    I'm not sure that says anything... About anything??

    I love regression analysis. A lot.
    I just can't comprehend why you didn't just run regressions on all the advanced single stats against each other instead of randomly multiplying three?

    If you wanted in all encompassing stat you could use PER... Or just one of the ones you mentioned ... I like VOR.


    Those advanced stats already have been tested yet still are obviously flawed in different ways. Why multiply them together?
     
    1 person likes this.
  18. LabMouse

    LabMouse Member

    Joined:
    Jun 27, 2008
    Messages:
    3,322
    Likes Received:
    79
    A great thread with an intellectual analysis. I do not know if people in TNT or ESPN even can understand your nice work, but this analysis just makes sense.
     
  19. RocketsJumer

    RocketsJumer Member

    Joined:
    Jan 31, 2013
    Messages:
    13,547
    Likes Received:
    1,441
    How do they define "top 5 players"? Is that starters or what the consensus 5 best players on the team? Most impactful?

    Warriors have been blessed with health. Don't want to wish anything on them but they are a Curry ankle away from being a moderately easy out.
     
  20. HeWhoIsLunchbox

    HeWhoIsLunchbox Contributing Member

    Joined:
    Oct 1, 2008
    Messages:
    3,393
    Likes Received:
    226
    <blockquote class="twitter-tweet" lang="en"><p><a href="https://twitter.com/theedmat">@theedmat</a> Average minutes played.</p>&mdash; HoopsHype (@hoopshype) <a href="https://twitter.com/hoopshype/status/580793010771902464">March 25, 2015</a></blockquote>
    <script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>
     
Loading...

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