1. Welcome to SportsJournalists.com, a friendly forum for discussing all things sports and journalism.

    Your voice is missing! You will need to register for a free account to get access to the following site features:
    • Reply to discussions and create your own threads.
    • Access to private conversations with other members.
    • Fewer ads.

    We hope to see you as a part of our community soon!

Odds of picking a perfect bracket

Discussion in 'Journalism topics only' started by biggerthanlashley, Mar 28, 2007.

  1. biggerthanlashley

    biggerthanlashley New Member

    This page caught my attention. Are the bracket odds on this page accurate? How could it be accurate, unless it's saying that if you picked every game completely at random, these are your odds? And if that's the case, why publish it, because nobody picks every game completely at random -- nobody puts the four No. 16 seeds in the Final Four, but if it's completely random, you're as likely to do that as you are to put the four No. 1 seeds in the Final Four.

    I'm not trying to pick on this designer or be an ass (I'm no anti-design Dyepack, I'm a regular poster using a pseudonym for my usual pseudonym), but I saw this page and the math issues bothered me and this site is good for a discussion of this kind of thing.

  2. Ace

    Ace Well-Known Member

    Hell with the bracket. Should I go for sainthood or supermodel?
  3. MilanWall

    MilanWall Member

    My mom once drew a royal flush playing video poker. I had no idea the odds were that low.
  4. three_bags_full

    three_bags_full Well-Known Member

    I drew a RF at a Friday night poker game about two months ago. Won a whopping 50 bucks.
  5. BTExpress

    BTExpress Well-Known Member

    2 to the 63rd power is indeed 9,223,372,036,854,775,807.

    But that in no way is an accurate measure of the odds of picking a perfect bracket.

    The paper assumes there is a 50-50 chance of picking the Florida-Jackson State winner correctly.

    And that is stupid to the 63rd power.

    Journalists, stay away from math.
  6. JR

    JR Well-Known Member

    Someone in authority needs to move this thread. :)
  7. John

    John Well-Known Member

    How do you calculate the odds on dating a supermodel?

    And I think I've seen four or five different numbers for the odds on picking a perfect bracket.

  8. That ideology is flawed, no doubt. But that is probably the closest represtentation to mathmatical odds to prediting it.

    Statistically speaking, every game has two possible outcomes - one team wins, one team loses. That's it. Even if Florida has no real chance of losing to a 16 seed, there's only two possible outcomes that can truly be measured. A win and a loss.

    Therefore this is the best representation.
  9. Ace

    Ace Well-Known Member

    Your chances=0

    Wealthy guy with access to drugs=100 percent
  10. blondebomber

    blondebomber Member

    Only if you fill out your bracket using a spread.

    Yes, there are only two possible outcomes, but without a handicap the likelihood of one over the other is nowhere near 50-50 in all but a few cases.
  11. Rhody31

    Rhody31 Well-Known Member

    It's not supposed to be 2 to the 63 power. I'm pretty sure it's supposed to be 63x62x61 ..... so on and so forth. At least that's what I remember from college. There are X amount of combinations and doing that gets all of them and thats how you figure out your chances.
  12. BTExpress

    BTExpress Well-Known Member

    A little research would give you better representation.

    Find the historical odds of a 16 seed winning, a 15 seed winning, a 14 seed winning, etc.

    And use those odds (instead of the 50-50 chance for every game) for at least the first round.

    Won't be perfect, but it the number will be a lot closer to reality than 1:9,223,372,036,854,775,807.

    The games are all independent of one another.

    So you (supposedly) have a 1:2 chance of picking each game right, 63 times.

    Now, if you wanted to calculate the odds of picking where every team would fall on the bracket before the bracket was published . . . the 64x63x62, etc., approach would be more suited.

    Here's an illustration:

    What are the odds of picking a "perfect" bracket this weekend?

    1 in 8 (2x2x2). There are only eight possible bracket combinations:

    UF over UCLA/UF over OSU
    UF over UCLA/UF over Gtown

    UCLA over UF/UCLA over OSU
    UCLA over UF/UCLA over GTown

    OSU over GTown, OSU over UF
    OSU over GTown, OSU over UCLA

    GTown over OSU, GTown over UF
    GTown over OSU, GTown over UCLA

    Doing it the 4x3x2x1 method would give you a 1:24 chance of picking the weekend bracket correctly. Which is way too high.
Draft saved Draft deleted

Share This Page