• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Warren Buffett offers $1 Billion.....

First, has it ever been done?

Second, not all permutations are equally possible. Someone with the relevant subject-matter expertise can probably narrow it down quite a bit.

Third, is it all possible permutations, or all the possible permutations after the initial brackets have been published by the NCAA (or whoever oversees March Madness)?
 
steve s said:
....to anyone who can fill out a perfect March Madness bracket. But before you get your hopes up, there's a quintillion possible permutations.

http://www.cnbc.com/id/101351399

Steve S
Click to expand...

Ah, a little advertising for Quicken on the shoulders of folks who can't do math. The odds are astronomical, even applying some basketball knowledge.

I'm also not wise to how BH works its insurance. If they're guaranteeing the pot, have they gone to the re-insurance market? Buffet may not be putting a billion of his own money on the table.
 
steve s said:
....to anyone who can fill out a perfect March Madness bracket. But before you get your hopes up, there's a quintillion possible permutations.

http://www.cnbc.com/id/101351399

Steve S
Click to expand...

For those of us who are basketball-impaired and not at all ashamed (I'm sure I'm not alone), what is this? Is it a prediction of the teams that will be playing in the tournament, or the individual players, or scores of games, or what?
 
Bob001 said:
For those of us who are basketball-impaired and not at all ashamed (I'm sure I'm not alone), what is this? Is it a prediction of the teams that will be playing in the tournament, or the individual players, or scores of games, or what?
Click to expand...

After the teams are picked and seeded into the tournament, filling out a bracket means picking the winner of each game.

For several years there were 64 teams (and therefore 63 games in the single-elimination tournament). There are now a couple of additional teams and and extra game or two. I don't know if those extra games are part of the bracket or if it starts after the field becomes 64.
 
Bob001 said:
For those of us who are basketball-impaired and not at all ashamed (I'm sure I'm not alone), what is this? Is it a prediction of the teams that will be playing in the tournament, or the individual players, or scores of games, or what?
Click to expand...

Once the teams are selected, they are fitted into a bracket of 64+/- (in recent years they have been adding play-in games) slots. The teams are seeded according to perceived strength (much like a tennis tournament) from 1-16 (there are actually four regions which compete separately until only one team remains from each region). In the first round, the #1 seed will play the #16 seed and the #2 seed will play the #15 seed, etc., on up to the #8 seed meeting the #9 seed. If all the seeds performed as predicted, then in the second round the #1 would play the #8, and in the third round it would be #1 against #4 and the regional finals would be the #1s against the #2s. There are a total of 63 games (32 first round, 16 second round, 8 third round, four regionals, two semi-finals and the championship game.)

Of course, things never work out that way. I don't think any #16 seed has ever won against a #1, but certainly some #15s have prevailed, and it is quite common for a #9 to prevail against a #8. So while it is not quite the same as predicting say, the outcome of 63 consecutive coin tosses, it is still vanishingly improbable.
 
theprestige said:
Third, is it all possible permutations, or all the possible permutations after the initial brackets have been published by the NCAA (or whoever oversees March Madness)?
Click to expand...

I'm not sure. They didn't mention that. But given the media's lack of literacy on scientific matters, I probably should have been a bit skeptical.

Steve S
 
Foolmewunz said:
Ah, a little advertising for Quicken on the shoulders of folks who can't do math. The odds are astronomical, even applying some basketball knowledge.
Click to expand...

How astronomical are the odds that if 100k basketball enthusiasts take this challenge, one will be right?
 
theprestige said:
First, has it ever been done?

Second, not all permutations are equally possible. Someone with the relevant subject-matter expertise can probably narrow it down quite a bit.

Third, is it all possible permutations, or all the possible permutations after the initial brackets have been published by the NCAA (or whoever oversees March Madness)?
Click to expand...

No, I don't believe it ever has.

I think I found that this year's tournament is 68 teams. If you just organize the bracket so each game has a team from the left and a team from the right, then you can assign a "left" or "right" winner to every game. So the possibilities are 2^67 or 147 quintillion.

Of course many permutations are unlikely. #1 seeds have won every first round game (although some were very close wins). Even the #2 seeds have only lost a handful of times. Let's say you take those 8 games off the table and assume no #1/#2 upsets in the first round. You can also assume that no team with a seeding higher than 12 plays in the final 3 rounds (has never happened before). I think you're still in the "quadrillions" for permutations.
 
Previous contests and results would indicate about 1 in 10,000,000 odds of getting the first round correct for the average guesser. Predictions should get more difficult after that, so we're talking worse than 1 in 100,000,000,000,000 odds even with applied knowlege. Even if 10 million people apply, I don't think he has much to worry about. Realistically priced reinsurance in that case should cost a few hundred dollars, so in reality whatever is the minimum amount for the insurer to cover overhead for insurance on a single event.
 
Charlie Wilkes said:
How astronomical are the odds that if 100k basketball enthusiasts take this challenge, one will be right?
Click to expand...

If you use "reduced" bracket (no upsets for #1/#2) then there are 10^59 permutations, many of which are horribly unlikely (like a final four of all high seeds). Lets be generous and guess that the number of "basketball likely" brackets is closer to 10^40. In that case, the odds of one winner would be

1 - (10^40 -1 / 10^40)^100k

That's about 1 in 10^-7 chances of a winner. I don't know how accurate my guess of 10^40 is. I think it's extremely generous.
 
BowlOfRed said:
If you use "reduced" bracket (no upsets for #1/#2) then there are 10^59 permutations, many of which are horribly unlikely (like a final four of all high seeds). Lets be generous and guess that the number of "basketball likely" brackets is closer to 10^40. In that case, the odds of one winner would be

1 - (10^40 -1 / 10^40)^100k

That's about 1 in 10^-7 chances of a winner. I don't know how accurate my guess of 10^40 is. I think it's extremely generous.
Click to expand...

OK, I am convinced. His money is safe.
 
Foolmewunz said:
Ah, a little advertising for Quicken on the shoulders of folks who can't do math. The odds are astronomical, even applying some basketball knowledge.

I'm also not wise to how BH works its insurance. If they're guaranteeing the pot, have they gone to the re-insurance market? Buffet may not be putting a billion of his own money on the table.
Click to expand...
BH is probably the largest player in the worldwide reinsurance market through their wholly owned subsidiary General Re. Many of the big deals are done by Ajit Jain, such as the $1B PepsiCo sweepstakes that they insured a few years ago. I think Buffett took a page from that play book for this event.
 
Last edited:
Charlie Wilkes said:
How astronomical are the odds that if 100k basketball enthusiasts take this challenge, one will be right?
Click to expand...

I think Bowl of Red has it higher than the number I'd read, but that was 63 games, not 65. For 63, it's a mere 9.3 quintillion to 1. (that's 9.3 billion billion).

And yes, there's some basketball and particularly historical basketball knowledge to it. Like the number of 16 seeds who have advanced so far - zero. And while 15s have made it through, especially Florida Gulf Coast in recent memory, they're not that common. But if you bet straight "favorites" based on the seedings I've read somewhere that there's never been a tournament that panned out accordingly.

While picking #1 and maybe even #2 in the first round is a pretty sure bet in most years, it gets closer to a coin toss as you go to 14 v 3, 13 v 4, and the infamous 12 v 5. (Infamous because there have been more 12s upsetting 5s than there have been 11s upsetting 6 an anomalous statistic that will likely iron itself out when there is a larger data set, as right now from '85 to '13, there have only been 112 such matchups.)

But, once you have all that figured out, of course you have to guess WHICH 12 is going to pull the upset. And you should likely pay no attention to the 1985 tournament. In the opening round, 2 number 11s, 1 number 12, and 1 number 13 all upset heavily favored opposition.

Does knowing all this make me any better at playing the pools? No. In a really good year I might get 45 out of 63 right, but have never hit 50. And that's playing two or more pools a year.
 
Foolmewunz said:
I think Bowl of Red has it higher than the number I'd read, but that was 63 games, not 65. For 63, it's a mere 9.3 quintillion to 1. (that's 9.3 billion billion).

And yes, there's some basketball and particularly historical basketball knowledge to it. Like the number of 16 seeds who have advanced so far - zero. And while 15s have made it through, especially Florida Gulf Coast in recent memory, they're not that common. But if you bet straight "favorites" based on the seedings I've read somewhere that there's never been a tournament that panned out accordingly.

While picking #1 and maybe even #2 in the first round is a pretty sure bet in most years, it gets closer to a coin toss as you go to 14 v 3, 13 v 4, and the infamous 12 v 5. (Infamous because there have been more 12s upsetting 5s than there have been 11s upsetting 6 an anomalous statistic that will likely iron itself out when there is a larger data set, as right now from '85 to '13, there have only been 112 such matchups.)

But, once you have all that figured out, of course you have to guess WHICH 12 is going to pull the upset. And you should likely pay no attention to the 1985 tournament. In the opening round, 2 number 11s, 1 number 12, and 1 number 13 all upset heavily favored opposition.

Does knowing all this make me any better at playing the pools? No. In a really good year I might get 45 out of 63 right, but have never hit 50. And that's playing two or more pools a year.
Click to expand...

Someone is going to have to drug the Gatorade to win Buffett's billion.
 
What odds would you get from the bookies for a similar prediction? Would it work as an accumulator or some other type of bet?
 
Foolmewunz said:
But if you bet straight "favorites" based on the seedings I've read somewhere that there's never been a tournament that panned out accordingly.
Click to expand...

It would be funny if it did this time though. One million people guess that way and get $1000 each.
 
Just a minor nitpick, but there are two definitions of a quintillion.

1. (Mathematics) (in Britain, France, and Germany) the number represented as one followed by 30 zeros (1030). US and Canadian word: nonillion
2. (Mathematics) (in the US and Canada) the number represented as one followed by 18 zeros (1018). Former Brit word: trillion
 
You must log in or register to reply here.

Back
Top Bottom