1. #1
    rfr3sh
    Blind Knucklehead
    rfr3sh's Avatar Become A Pro!
    Join Date: 11-07-09
    Posts: 10,229
    Betpoints: 604

    Modelling NBA Question

    What data do you find more relevant, the last 5 game stats , season stats or a combination of both?

    been trying to figure this out for a while it is tough to say

  2. #2
    samserif
    samserif's Avatar Become A Pro!
    Join Date: 09-19-11
    Posts: 63
    Betpoints: 165

    Not trying to be discouraging, but this looks like the sort of thing you'll need to figure out on your own. Even when someone posts this kind of information on here, I'm skeptical until I check it myself. Besides, if you discover that the last five games works well, you'll then wonder if six is better... or maybe four, or ten, or...

  3. #3
    rfr3sh
    Blind Knucklehead
    rfr3sh's Avatar Become A Pro!
    Join Date: 11-07-09
    Posts: 10,229
    Betpoints: 604

    my model shows both sets of data but hard to say which is more important my guess is the correct number is somewhere in the middle with some sort of weighting

    this is for totals btw

  4. #4
    JohnAnthony
    JohnAnthony's Avatar Become A Pro!
    Join Date: 04-30-09
    Posts: 5,110
    Betpoints: 12

    Quote Originally Posted by rfr3sh View Post
    my model shows both sets of data but hard to say which is more important my guess is the correct number is somewhere in the middle with some sort of weighting

    this is for totals btw
    Like samserif said, anyone with good indicators on where to put more weight would probably keep the information to himself. People here are very helpful in regards of methods of building and etc...

    But if it gets you any further, I also tackle with two indicators mostly, and most of the time I'm left wondering which one to put more weight on when making a pick (if both are showing a GO, in terms of making the pick, it's a sizing issue).

    Good luck,

  5. #5
    Abelesw
    Abelesw's Avatar Become A Pro!
    Join Date: 12-03-11
    Posts: 11
    Betpoints: 56

    I agree with JohnAnthony. Finding the most accurate weighting method is an exact science, I'll only pull the trigger if both are showing a go.

  6. #6
    JohnAnthony
    JohnAnthony's Avatar Become A Pro!
    Join Date: 04-30-09
    Posts: 5,110
    Betpoints: 12

    Quote Originally Posted by Abelesw View Post
    I agree with JohnAnthony. Finding the most accurate weighting method is an exact science, I'll only pull the trigger if both are showing a go.
    In my case (and remember I'm no Think-Tank master, not even close), I'd also bet if one indicator is showing a strong GO, and the second is slightly positive. In this occasion, I put more weight on H2H, and try to see who is going to be the "dominating" factor in deciding the result. The play is also usually a lot smaller in this case.

    That's it, I'm going to shut up now

  7. #7
    rfr3sh
    Blind Knucklehead
    rfr3sh's Avatar Become A Pro!
    Join Date: 11-07-09
    Posts: 10,229
    Betpoints: 604

    yeah you guys do pretty much what ive been doing im just always looking to get better maybe after this season I can do back testing if I have time

  8. #8
    MonkeyF0cker
    Update your status
    MonkeyF0cker's Avatar Become A Pro!
    Join Date: 06-12-07
    Posts: 12,144
    Betpoints: 1127

    So when a team has played their last (whatever arbitrary number you're trying to put here) games against the likes of the Wizards, Raptors, Pistons, Hornets, and Bobcats, you should preload it to weigh more than another arbitrarily-lengthed set of games against the Bulls, Magic, Thunder, Nuggets, and Heat simply because the former set occurred last? Is that what you're attempting to do?

  9. #9
    rfr3sh
    Blind Knucklehead
    rfr3sh's Avatar Become A Pro!
    Join Date: 11-07-09
    Posts: 10,229
    Betpoints: 604

    Yeah I was thinking about adding in the sos too, but from what you are seeing it seems that last 5,6,7 whatever is not really relevant

Top