How much would a bracket be worth?

[albionmoonlight]

I was looking at a completed bracket this morning and thought, "Wow, how much would this information have been worth just one month ago?"

And then I thought, "really, how much would this be worth?" Even assuming that a Vegas whale got his hands on it, there would be a limit to how much he could bet without attracting suspicion and/or making so much news as to affect the outcome of the games.

And no one entity, I think, would be willing to put itself on the hook for billions of dollars for one event. At the same time, someone who knew what they were doing could enter a lot of high stakes pools and rack up a lot of cash.

So, my question (and I have not yet decided if it is interesting or not)--how much could someone with a healthy set of connetions in the gambling world have made with complete foreknowledge of the NCAA pool?

[KevinNU7]

No one has those connections.

[sooner333]

What about the Sports Alamanac in Back to the Future II???

[gottimd]

Biff Tanner did it and made loads of money.

:D Posted the same time Sooner did.

[albionmoonlight]

No one has those connections.
Well, one question would be--how much could "you" have made in the pool, but I imagine that that is more of a function of cash on hand for people than anything.

I realize, however, that the answer to the question depends so much on the details of this idealized gambler that the question may just not have an answer. Which I why I don't even know if it is interesting.

There still does seem, though, to be a limit. I can't picture someone making, say, 2 billion dollars with NCAA pool wagers, no matter how much he had to bet and how well connected he was.

[Ksyrup]

What about the Sports Alamanac in Back to the Future II???
My thoughts exactly. Had I been involved in the real-life Back to the Future, the movie would have ended once I found that book. In fact, thanks to that movie, if I ever travel in time (either way), my first stop is to a gambling establishment.

I'm not a gambler, but then, with this kind of information, it's not really gambling...

[Maple Leafs]

There have been (and may still be) contests where a 100% correct bracket selection triggers a ridiculous prize -- I think it may have been as high as a billion dollars.

[JeeberD]

I think in the past ESPN has offered $1 million to anyone who had a perfect bracket in the ESPN pools, and if I remember correctly someone actually did it last year.

I could certainly be wrong, though...

[gottimd]

There have been (and may still be) contests where a 100% correct bracket selection triggers a ridiculous prize -- I think it may have been as high as a billion dollars.

I bet you a chick wins it at some point. And they will ask her if she watches basketball, and her answer will be no. She just chose who had a better mascot, better colors, a better area to live, etc.

[Maple Leafs]

I think in the past ESPN has offered $1 million to anyone who had a perfect bracket in the ESPN pools, and if I remember correctly someone actually did it last year.If true, I didn't hear about it. I strongly suspect the contest will never be won.

And (from a google search) in case you're wondering why people don't just fill out and submit every possible combination:
What are the odds of getting a perfect bracket??

1 out of 2^64, or 18,446,744,073,709,551,615 to one against

And if you were to fill out one bracket per page on newsprint (lightweight paper, 0.003" thick) it would form a stack that would reach from here to Pluto and back sixty times!

If each person on earth took exactly one minute to fill out a bracket, it would take a little under 5,850 years to fill out all possibilities.

There are nearly 1.8 million different brackets for every dollar in the US Gross Domestic Product.

[gottimd]

If true, I didn't hear about it. I strongly suspect the contest will never be won.

And (from a google search) in case you're wondering why people don't just fill out and submit every possible combination:

In "Real Genius", the character Laslow has a machine that fills out stuff automatically. I'm sure there is technology that can be written/programmed to fill out every possible combo in a decent amount of time. Then it comes down to paper source....just use your program to fill out the internet ones. Then I guess it comes down to the money you have to put in to fill out and submit all of those brackets!

[Wolfpack]

You can reduce that stack by some measure by making educated guesses in the first two rounds of the tournament (there isn't any other way to do it and still live). Once you have your Sweet 16 guess, you can then spend 45 days (assuming one minute per bracket combination) putting in all the combinations at that point. Of course, once you are done, the tournament will have long since been decided.

[albionmoonlight]

In "Real Genius", the character Laslow has a machine that fills out stuff automatically. I'm sure there is technology that can be written/programmed to fill out every possible combo in a decent amount of time. Then it comes down to paper source....just use your program to fill out the internet ones. Then I guess it comes down to the money you have to put in to fill out and submit all of those brackets!
If you assume that a program could fill out and submit a bracket online every second (and that the site putting itself at risk would allow itself to be spammed in that manner), then it would still take over 584,942,417,355 years to submit every possible bracket.

[Passacaglia]

Aren't the odds actually 1 in 2^63 anyway? That cuts it in half.

I'd think the completed bracket would be worth more on a game-by-game basis, if you're worried about triggering suspicion.

This reminded me of an extra credit assignment I gave out last week. I gave my students an old AP exam, and told them I didn't care how well they did on it, and I'd gave them the same amount of extra credit just for trying it. One student asked me if I would give more extra credit if he get EVERY SINGLE QUESTION right. :rolleyes:

[bigphesta]

There are 65 teams in the tourney - remember the play in game?

[larrymcg421]

Obviously if you had such foreknowledge (whether by time travel or any other method), the thing to do would be to bet on a game by game basis and purposefully miss a couple games. For instance, just because you know Bucknell is gonna beat Kansas, you might not want to bet that one.

[gottimd]

Obviously if you had such foreknowledge (whether by time travel or any other method), the thing to do would be to bet on a game by game basis and purposefully miss a couple games. For instance, just because you know Bucknell is gonna beat Kansas, you might not want to bet that one.


I would rather get the longshot bet in (better odds payout) and miss the Pick em's.

[Huckleberry]

To better avoid suspicion, I would think a good strategy would be to simply bet the same amount on the money line in each game. Making sure to hit most of the top possible payouts, but not all.

Bet on UAB, Wisconsin-Milwaukee, and Bucknell, but bet on Syracuse at the same time.

[Glengoyne]

We were too swamped to do the office pool this year. I didn't even get my bracket filled out until about noon Pacific time on the opening Thursday. I would so have won the office pool again this year. I've only lost once since I started here, that was when Syracuse won it all. Since I almost always expect them to drop early like they did this year, I didn't do all that well that year. This year, not only did I have NC over the Illini for the final, but I only missed 13 winners for the whole bracket.

[gottimd]

To better avoid suspicion, I would think a good strategy would be to simply bet the same amount on the money line in each game. Making sure to hit most of the top possible payouts, but not all.

Bet on UAB, Wisconsin-Milwaukee, and Bucknell, but bet on Syracuse at the same time.

You think a 50 game parlay hitting would raise suspicion?

[Passacaglia]

Damn play in game!

[st.cronin]

You think a 50 game parlay hitting would raise suspicion?

Not if you find some ditzy secretary type to place your bet.

[sterlingice]

I would rather get the longshot bet in (better odds payout) and miss the Pick em's.
Yeah, maybe botch that Arizona/Okie State and Louisville/Ga Tech game ("Oops... I figured it would be a close win for Tech"). Those types of game and put some money down on Bucknell and Vermont because you "like the matchups" and put down a longshot UMW in the Sweet 16 because then you can bet on both games under the guise that you had them in your bracket. And get an MSU grad to put some sweet cash on them to make the Final Four.

Speaking of which, you cut in a bunch of friends, provided they believe your crazy time travel story ;) Like those groups which went around trying to win lotteries. That way there's less suspicion and it's more spread around. You could be co-workers in that you used to work together and after the tourney, you'll all not be working :)

SI

[sterlingice]

Not if you find some ditzy secretary type to place your bet.
Yeah, one of those "I pick based on mascots" type.

SI

[primelord]

If true, I didn't hear about it. I strongly suspect the contest will never be won.

And (from a google search) in case you're wondering why people don't just fill out and submit every possible combination:
What are the odds of getting a perfect bracket??

1 out of 2^64, or 18,446,744,073,709,551,615 to one against

And if you were to fill out one bracket per page on newsprint (lightweight paper, 0.003" thick) it would form a stack that would reach from here to Pluto and back sixty times!

If each person on earth took exactly one minute to fill out a bracket, it would take a little under 5,850 years to fill out all possibilities.

There are nearly 1.8 million different brackets for every dollar in the US Gross Domestic Product.
The 1 out of 2^64 is the odds of randomly guessing a perfect bracket. However we all know that the tournament is not a random series of events. That numbers includes the posibilities of 4 16 seed teams making the final 4. That will obviously never happen. Just using sound judgement in picking your teams would improve your chances of hitting a perfect bracket exponentially.

Picking a perfect bracket would certainly be a difficult thing to do, but nowhere near as difficult as that number suggests.

[panerd]

The 1 out of 2^64 is the odds of randomly guessing a perfect bracket. However we all know that the tournament is not a random series of events. That numbers includes the posibilities of 4 16 seed teams making the final 4. That will obviously never happen. Just using sound judgement in picking your teams would improve your chances of hitting a perfect bracket exponentially.

Picking a perfect bracket would certainly be a difficult thing to do, but nowhere near as difficult as that number suggests.

Actually I think it is almost more difficult considering what you (and others) have said. Where do you stop your assumptions? That the 1 seeds will win the first round? That removes how many possibilities? A couple thousand? Because if you go any farthur the nonsense scenerios have happened plenty. (#2's and 3's losing in the first round, #1's losing in the second, #12's making the elite eight)

[Airhog]

Man If I could go back in time I would screw gambling. I would just get the pick 6 numbers the week before some big jackpot hit, and score there. Much easier and less suspicion

[primelord]

Actually I think it is almost more difficult considering what you (and others) have said. Where do you stop your assumptions? That the 1 seeds will win the first round? That removes how many possibilities? A couple thousand? Because if you go any farthur the nonsense scenerios have happened plenty. (#2's and 3's losing in the first round, #1's losing in the second, #12's making the elite eight)
Well there is no way it is more difficult than just randomly picking the teams. My point was not to say that it would be easy or even likely ever to happen, but rather that in terms of pure odds it is not anywhere near as difficult as that number suggested.

In the 1st round of this years tournament 75% of the games were won by the higher ranked seed. I follow college basketball, but never to the extreme point where I can accurately judge the matchup of any two teams. So in general outside of Big 12 teams and the really top teams I tend to pick for the most part the high seeds in the 1st round and somewhat randomly choose some upsets. Occasionally I will make an educated predicition such as Kansas losing to Bucknell. Whenit is all said and doen I would say my 1st round is generally right around 75% favorites and 25% upsets. It is FAR easier to guess 8 upset games than 32 like the original number is assuming.

In the 2nd round this year the higher ranked seeds won 81% of the games. Again other than going with just the higher ranked seeds you only had to guess 3 upsets correctly. Once you get into the sweet 16 the cindarella stories tend to die off and the other teams left are geenrally top teams and it is much easier to start making educated decisions on who will win etc. For example Louisville over Washington was an easy upset pick this year etc.

Now with all the being said I think you would have to be VERY lucky to pick a perfect bracket, but again my point was just that it isn't a random process and is not anywhere near as unlikely as that article made it seem.

[gottimd]

I would agree with primelord in regards to 1st round picking. In running the tournament pool for a few years, I keep records for Hall of Fame purposes, and taking a sample pool of 450 people, the average in the past 5 years has been 24.2 games correctly picked in the first round. That is roughly 75%. 29 is the highest I have seen picked correct in the 1st round in running a pool.

[ColtCrazy]

Someone will eventually make a program that will automatically fill them out, and if they could put in some restrictions to the program, that would save time
Like
No No. 16 or 15 wins a game (I know some 15s have, but it's so rare it's probably worth the risk not to have them)
No 13 see or lower in the regionals
No 9 seed or lower in the final four.

That would take out lots of options right there.

[Maple Leafs]

But you guys are moving from "how hard would it be to guarantee a correct bracket" to "how hard would it be to guarantee having a pretty good chance at a correct bracket". Big difference.

You can't throw out 15 seeds, for example, because they do win. Sure it doesn't happen often, but imagine spending countless hours and money to enter a contest and then losing because of a 15 seed?

I could accept the #16 seed restriction since it's never happened, but even that's risky if you want a guaranteed win.

Beyond the math, the fact that some people spend an amazing amount of time and resources on this sort of thing and yet nobody in the history of the world has ever had a perfect selection should give you an idea how unlikely it is.

[QuikSand]

Given the drift of this topic...

Some years ago, there was a syndicate (Australians, I think?) who decided that there was an opportunity for positive expectations from state lotteries under certain circumstances. However, they needed to buy literally millions of tickets to cover all the potential outcomes. They actually did this once, at least -- they went to lottery outlets with boxes of pre-prepared lottery forms, and occupied the outlets for hours on end, to try to buy the millions of tickets necessary. As I recall, (this was in the early 1990s, I'd say) they did not manage to cover every single number combination, but they did get lucky and win the prize.


Just an interesting diversion story, I suppose. But I'm with Maple Leafs -- those who are suggesting that a perfect bracket is practically possible are missing the sheer magnitude of the numbers involved here. Even if you make some reasonable assumptions about likelier outcomes, it's extraordinarily unlikely to ever happen, by anyone.

[Blade6119]

But all this talk of suspicion, what would you think if you were vegas....would you really ever believe this guy came from the future, or any other way to know the games...flat out, i dont think you would get in trouble. You would be famous for doing it, or rich and not famous, but what would they do..come after you for guessing right. There is no way i can think of to cheat the bracket in adavance unless your from the future...and come on, even vegas wouldnt buy that story.

[primelord]

Just an interesting diversion story, I suppose. But I'm with Maple Leafs -- those who are suggesting that a perfect bracket is practically possible are missing the sheer magnitude of the numbers involved here. Even if you make some reasonable assumptions about likelier outcomes, it's extraordinarily unlikely to ever happen, by anyone.
I don't disagree with that. I guess I just have an issue with saying the odds of picking a perfect bracket are 18,446,744,073,709,551,615 to 1 when they are in fact significantly better than that. It is possible to have a significantly better chance of picking a perfect bracket than the number listed and still have virtually no shot of accomplishing it.

Just eliminating the 16 seeds alone eliminates an enormous number of bracket combinations. I don't see the point of listing an exact number in article when that number isn't correct. Clearly he was going for shock value and there is no real way to determine the exact odds when someone is making "informed" choices, but I just think if you are going to present the odds of something happening not only should the be accurate, but they shouldn't be way off like that one was.

[mckerney]

Given the drift of this topic...

Some years ago, there was a syndicate (Australians, I think?) who decided that there was an opportunity for positive expectations from state lotteries under certain circumstances. However, they needed to buy literally millions of tickets to cover all the potential outcomes. They actually did this once, at least -- they went to lottery outlets with boxes of pre-prepared lottery forms, and occupied the outlets for hours on end, to try to buy the millions of tickets necessary. As I recall, (this was in the early 1990s, I'd say) they did not manage to cover every single number combination, but they did get lucky and win the prize.


Just an interesting diversion story, I suppose. But I'm with Maple Leafs -- those who are suggesting that a perfect bracket is practically possible are missing the sheer magnitude of the numbers involved here. Even if you make some reasonable assumptions about likelier outcomes, it's extraordinarily unlikely to ever happen, by anyone.

There was a story last year that I think took place in Missouri that involved a pick 3 game and a promotion that they were doing that involved a bonus drawing on a random day of the week. When that hadn't come by thursday, the people knew it had to come on Friday, and two winning tickets would bring in more money than it would cost to buy ever single combination of numbers. So the guy asked everyone he knew if they were interested and bought several full sets of numbers, though I don't remember how much he wound up winning.

[primelord]

Actually I think it is almost more difficult considering what you (and others) have said. Where do you stop your assumptions? That the 1 seeds will win the first round? That removes how many possibilities? A couple thousand? Because if you go any farthur the nonsense scenerios have happened plenty. (#2's and 3's losing in the first round, #1's losing in the second, #12's making the elite eight)
If you only made the assumption that the #1 seeds would win their first round games and you picked every otehr game at random your odds of a perfect bracket go from:

18,446,744,073,709,600,000 to 1

to

1,152,921,504,606,850,000 to 1

That's a difference of over 17 quintillion. Not quite just a couple thousand. :)

It would also be very reasonable to eliminate the 15 seeds. They have only ever won 4 times. Removing the 15 and 16 seeds drops your odds to:

72,057,594,037,927,900 to 1

Now obviosuly the number is going to be even smaller than that because most of the games in the first couple rounds are going to come out as expected with the top seed winning.

So when it is all said and done is it still virtually impossible to pick a perfect bracket? Yes. Is it substantially more likely than the odds given in that article? Yes.
<table x:str="" style="border-collapse: collapse; width: 137pt;" border="0" cellpadding="0" cellspacing="0" width="183"> <col style="width: 137pt;" width="183"><col> <tbody><tr><td class="xl24" style="height: 12.75pt; width: 137pt;" x:num="1.8446744073709601E+19" align="right" height="17" width="183">18,446,744,073,709,600,000
</td></tr></tbody><tr style="height: 12.75pt;" height="17"></tr> </table> <table x:str="" style="border-collapse: collapse; width: 137pt;" border="0" cellpadding="0" cellspacing="0" width="183"> <col style="width: 137pt;" width="183"><col> <tbody><tr><td class="xl22" style="height: 12.75pt; width: 137pt;" x:num="1.8446744073709601E+19" align="right" height="17" width="183">18,446,744,073,709,600,000
</td></tr></tbody><tr style="height: 12.75pt;" height="17"></tr> </table>

[QuikSand]

If you only made the assumption that the #1 seeds would win their first round games and you picked every otehr game at random your odds of a perfect bracket go from:

18,446,744,073,709,600,000 to 1

to

1,152,921,504,606,850,000 to 1

That's a difference of over 17 quintillion. Not quite just a couple thousand.

Okay, so you reduce the stack of entires 16-fold. So your stack of entries would only have to go all the way to Pluto and back four times instead of 60 times. Sounds great. Call me when you have them all filled out.

You're right that making this assumption does indeed wipe out the majority of the possible options... but to suggest that this movement is that big a deal is a little misleading, I think.


I see your point... but to suggest that this is somehow a thoroughly manageable feat goes too far, I think.

[gottimd]

Okay, so you reduce the stack of entires 16-fold. So your stack of entries would only have to go all the way to Pluto and back four times instead of 60 times. Sounds great. Call me when you have them all filled out.

Done! Whats your number?

[QuikSand]

Done! Whats your number?

It's printed on the stickers on the back of trucks all across the country... should be easy enough to remember.

[gottimd]

It's printed on the stickers on the back of trucks all across the country... should be easy enough to remember.

Rather than call....how is your driving?

[albionmoonlight]

Or, to put it another way. Let's say that I were to tell you how the entire first round went, and you then just had to fill out brackets to cover the remaning 31 games. Filling out a bracket a second, it would still take you 68 years to do it.

[primelord]

Okay, so you reduce the stack of entires 16-fold. So your stack of entries would only have to go all the way to Pluto and back four times instead of 60 times. Sounds great. Call me when you have them all filled out.

You're right that making this assumption does indeed wipe out the majority of the possible options... but to suggest that this movement is that big a deal is a little misleading, I think.


I see your point... but to suggest that this is somehow a thoroughly manageable feat goes too far, I think.
Again that wasn't the point. I never said this was something that was likely. I just said that the odds were significantly lower than what was being presented. And the odds are FAR lower than even what I presented. I was simply demonstrating the massive difference in just assuming the #1 seeds would win their game.

I also never suggested you could even fill out all the likely brackets. I just didn't see the point of the article listing incorrect odds of accomplishing something. If the numbers aren't right I don't think they should be listed.

[QuikSand]

"Significantly lower" is probaly the issue at point. You're right, in a mathematical sense, that making reasonable assumptions might pare down the 18 quintillion significantly to something like only one quadrillion is a big step. However, in any kind of practical terms, even a quadrillion (or a thousanth of that) is a number so far beyond anything we could actually accomplish that the difference could probably still be considered to be insignificant.


Hope that is clearer. I don't think we really disagree on this. I see the post you were reacting to, and you are correct in pointing out its error.

[gottimd]

So this has gone from a "If you went back in time" thread, to a "do you really think you can fill out that many".

Selection comes out on Sunday with games shortly after that, so unless the volume of output can be possible within a few days or so, even with machine, then the arguments are pointless.

However, you could join a pool, act like the expert and gaurantee the wrong winners, and maybe persuade others to take the wrong teams, as well as gamble on the longshots, bet some chicks some wild sex acts, or whatever if you had the "Sports Almanac from the future".

But in terms of filling out paper work to Pluto, don't think you will hit that goal in a matter of days unless you took one of those speed writing seminars, the really really really advanced lecture of it.

[Maple Leafs]

So this has gone from a "If you went back in time" thread, to a "do you really think you can fill out that many".Pretty much. That tends to happen around here when math is involved.

(But I stand by my original point -- If I somehow knew all the outcomes, I'd forget Vegas and find as many "perfect bracket" contests as I could.)

[Daimyo]

Even if you used a computer to generate all possible outcomes, skipped the paper printouts, and did it all online it would take approximately 2,700,000,000 years to upload the info over a standard DSL connection (assuming 9 characters/team name transmitted in pure text). Or I guess you could burn them onto 2,300,000,000,000 DVDs and deliver them that way... *shurg*

[Daimyo]

DOLA, whoops... I didn't realize this was a dead thread. Interesting to me anyway because its so similar to the discussions of password length and character set depth vs time to brute force crack...

[primelord]

DOLA, whoops... I didn't realize this was a dead thread. Interesting to me anyway because its so similar to the discussions of password length and character set depth vs time to brute force crack...
That's what distributed computing is for. :)