Let's Play A Game With On/Off Court Stats

Discussion in 'Butler Basketball' started by Jared Grubbs, Nov 6, 2018.

  1. Jared Grubbs

    Jared Grubbs Administrator Staff Member

    Money:
    $15,957
    I know there’s sarcasm there, but I’m going to answer that literally, as a lot of people reading get confused by these. It’s not really something that can be used mid game like that. I would hope he uses similar data later in the season as we get a larger sample size against better teams, but even then it’s still just one piece of data. I like looking at it because it’s the only data we have that factors in anything outside the box score. That doesn’t make it better, if just gives us a different angle to look at.


    Sent from my iPhone using Butler Hoops mobile app
     
    fdr, bwesson15, harvesting and 2 others like this.
  2. Hinkle

    Hinkle Well-Known Member

    Money:
    $7,053
    Nate Fowler casually continuing his status as the most confounding Butler player ever and the most interesting case study for advanced analytics in basketball.


    Sent from my iPhone using Butler Hoops mobile app
     
  3. bumba

    bumba Well-Known Member

    Money:
    $7,476
    The most overanalyzed 6 points & 5 rebounds per game ever.
     
  4. Hinkle

    Hinkle Well-Known Member

    Money:
    $7,053
    Are you really not interested in what the *team* is doing when he’s on the floor, and why? If you are a fan of the team more than the players, I don’t know how that’s possible.


    Sent from my iPhone using Butler Hoops mobile app
     
  5. bumba

    bumba Well-Known Member

    Money:
    $7,476
    I don't need a calculator to recognize Joey Brunk is playing better.
     
  6. Insane Dawg

    Insane Dawg Well-Known Member

    Money:
    $12,653
    No sarcasm intended. Nor was I trying to be facetious. The numbers you shared alone proves Fowler is the most valuable or important player on the team yet he is averaging only 17 minutes thru the first three games. Based on that, even though he is a starter, he is seventh in minutes played. What I was getting at is LJ's substitutions do not reflect the numbers you posted as to his value. Either LJ isn't looking at these numbers after each game or there is something else going on.
     
  7. Hinkle

    Hinkle Well-Known Member

    Money:
    $7,053
    He may be! But the fact is the team has done better on both ends when Fowler is out there. The only explanations for this are: (A) Fowler is better and your eyes deceive you; or (B) something about the team strategy or lineup or periods of the game during which Brunk plays leads the team to play worse when he plays, despite being as good or better of a player.

    *I don’t know which of those it is.* My hunch is that he’s just a better player, but he’s better in unconventional ways so it doesn’t seem like it. But I admit I don’t *know* that. (And you absolutely don’t know that Brunk is better. You’re indeed fighting the uphill battle to support your position because the simpler explanation is that Fowler is just better.) But it’s these kinds of questions and insights that make the data so valuable.


    Sent from my iPhone using Butler Hoops mobile app
     
  8. the_speakers_lab

    the_speakers_lab Well-Known Member

    Money:
    $4,241
    SSS
     
  9. bumba

    bumba Well-Known Member

    Money:
    $7,476
    I work with market research every day. Every. Day.

    Data can tell you anything you want - there is no magic bullet hidden inside numbers. That's why I personally find this conversation tiring. For years people have used metrics to justify performance instead of actual performance.

    I think Nate is a great ambassador of BU, has been recognized as a fantastic student & is more than likely a great kid. Brandon took a chance on size because we couldn't sign a 5 after Smith graduated & he was from the same school as Monserez and Barlow.
     
    estar20dawg likes this.
  10. Hinkle

    Hinkle Well-Known Member

    Money:
    $7,053
    The team objectively has performed better when Nate is on the floor (unless you think scoring more points on offense is not “better” and allowing fewer points on defense is not “better”). It performed best when he was on the floor last year too. Those are objective facts. This is not subject to interpretation. Period.

    What is subject to interpretation is *why*.


    Sent from my iPhone using Butler Hoops mobile app
     
  11. Insane Dawg

    Insane Dawg Well-Known Member

    Money:
    $12,653
    Hinkle I agree with what you just said. If the numbers say Fowler is the most positive influence on the floor then it is what it is. The question as to why the team performs better when he is on the floor versus anyone else is an interesting question. To me if he has that much positive influence on the team he should play more that seventeen minutes a game. Or there may be variables that aren't being measured.
     
    Hinkle likes this.
  12. I Am Butler

    I Am Butler Well-Known Member

    Money:
    $4,119
    I find this type of information as fascinating as anyone, but the problem with this data right now is there isnt enough of it. You can just glance at how close all of the offensive PPP are and see the sample isnt big enough to reliably differentiate performance.

    What I did notice against Ole Miss is that their guards were getting a lot of easy layups when Brunk was in the game. Not sure how much of that to put on Brunk or Fowler, as are guards were just getting beat, and maybe it just happened to be when Brunk was in the game.

    I'm still not really sure what to make of the numbers overall. The early performance of Brunk is maybe the most staggering example of the stats not matching the eye test. Nobody watching the games is thinking, "Damn, Fowler is just the superior player." Why is that?
     
    Hinkle likes this.
  13. I Am Butler

    I Am Butler Well-Known Member

    Money:
    $4,119
    This is what I keep coming back to. Would love to hear somebody with a deep understanding of these numbers speak to this.

    Advanced basketball analytics still lag far behind sports like baseball and there just isnt as much good insight on some of these metrics.

    How do you remove a player that uses such a low number of possessions to score his points and the team performs worse offensively. This was the conversation with Wideman and th early returns say the same thing about Brunk. Its counterintuitive. What is happening that makes this true?
     
    fdr and Hinkle like this.
  14. DawgsMD

    DawgsMD Forum Psychiatrist VIP Member

    Money:
    $2,294
  15. BoyGeoff

    BoyGeoff Active Member VIP Member

    Money:
    $202
    I’m just glad to see that as a combination, that are playing very well at the center position. Before the season my biggest question was how will our bigs play. So far they have played very well. Following each of the 1st 3 games I said to myself: “if they had our bigs, it might have been a different outcome.” Detroit and Ole Miss may have gone the other way. I guess what I am saying is, our bigs have outplayed the competition in each of the 3 games we have played. So if they continue to get better each game, we have other fish to fry. I like having two that can play well.


    Sent from my iPhone using Butler Hoops
     
    estar20dawg likes this.
  16. Hinkle

    Hinkle Well-Known Member

    Money:
    $7,053
    This is a good point generally but the reason I’m giving it more weight than a three game sample deserves is that he led it last year too. So it’s essentially a 38 game sample that says the team performs really really well with Fowler on the floor. Maybe with more games/a larger sample Brunk will catch up to him, as the sample size with Brunk is indeed extremely small.



    Sent from my iPhone using Butler Hoops mobile app
     
    bwesson15 and I Am Butler like this.
  17. dawgs2014

    dawgs2014 Well-Known Member

    Money:
    $2,589
    Who said that good players abilities have to be skill based?

    Good players contribute to maximizing net margin, right? That’s exactly what’s measured. Who cares how they get there. It’s not predictive, but it’s valuable.


    Sent from my iPhone using Butler Hoops
     
  18. I Am Butler

    I Am Butler Well-Known Member

    Money:
    $4,119
    I never said they had to be skill based. I'm specifically pointing out that these numbers DON'T lean heavily on traditional skill based metrics. What I'm trying to understand is exactly why some players seem to have such a large effect (positive or negative) on the team despite not necessarily having physical skills that match. I dont profess to having this answer because I frankly dont have enough understanding of advanced basketball analytics. In fact, because of where the basketball industry is at with their development of usable analytics, I'd say very few people can truly answer this question.

    I most definitely disagree with the notion that the numbers are valuable if they aren't predictive. That idea flies in the face of the sabermetrics industry. I believe these numbers do have some predictive value, but understanding how to use it is key. It's obvious you can't just take the top 5 players in these rankings and make them your primary rotation. That's why I'd like to know if and how the coaching staff relies on this feedback.
     
    Last edited: Nov 20, 2018
    Insane Dawg likes this.
  19. dawgs2014

    dawgs2014 Well-Known Member

    Money:
    $2,589
    I think that a primary focus of analytics to be to raise questions. Why do these numbers reflect this way? Let’s go look at the tape.
    Any stats algorithm that can calculate these numbers can calculate more in depth analysis of the variance with not much more work (ie the types of shots change, rebounding percentages change etc) and those can provide more evidence of where to look.
    Relying on stats by themselves or on tape by itself is equally dangerous because neither show the whole picture


    Sent from my iPhone using Butler Hoops
     
  20. troggy

    troggy Active Member

    Money:
    $596
    Probably because the larger difference in the numbers is on the defensive end, which isn't as noticeable. As has been mentioned several times, it's a very small sample size on Brunk, so I wouldn't read that much into it yet. Just be happy that Brunk has emerged and is going to be a major player on this year's team.
     
    dawgs2014 likes this.

Share This Page