A different idea: We tell you where to guess, and YOU submit the winning answer for us.
... (confused pause) ...
But how will we get our share
of the prize?
... (confused pause) ...
What if I just take the money?
That's the whole point! We want you to win $40,000 and not give us a dime*.
Do you notice one thing common to all these "Red Balloon" strategies?
You don't get a dime unless you are the first to spot the balloon.
So if a balloon is clearly visible from the highway on your way to work, you're very unlikely
to win anything. Of course, all the teams will try to hide this from you.
The real problem is that most organizers won't know what to do with all the data they receive.
There will be wrong submissions, whether intentional or accidental.
If the small group of people "leading" a team gets overwhelmed and can't piece
together all the information, the whole team comes crumbling down and there is
zero chance of anybody on the team winning.
How do you choose 10 locations from a mess of contradictory submissions? It's not
that easy.
Maybe by looking for peaks in a kernel density estimator.
Or perhaps with some sort of
Gaussian mixture model (yes, we are experts at this kind of thing and have thought
through lots of possibilities).
From even 25 close guesses, though, there are
over three million ways
to pick 10 locations.
Remember, if you're off on ANY balloon by only a mile, the
full submission doesn't count! We've been working hard on strategies for
aggregating information, but we still think it is too hard to choose just 10 balloon location
guesses.
Other teams can wait for more data to accumulate to get more accurate estimates,
but only the first team to submit the correct answer wins. If you wait, you lose.
What do we stand to gain?
We mostly just want the recognition. We would get way more than $40k worth
of benefit from lots of press saying
that we can deliver creative, practical solutions to hard problems like this.
If you win because of us, we hope you'll tell the press that we helped. We figure
it's a small price for you to pay for the $40,000 you'll be pocketing.
Think of it like an affiliate marketing deal (we'll reveal ourselves later).
Also, we're putting some ads on the page and hope to make a small, quick buck.
We'd try to win flat out if we could, but we really think
it would be too hard and slow to narrow it down to only 10 locations.
Finally, we're not storing any email addresses or dealing with any fancy
legal issues, so there's much less effort required on our part. We built
the underlying statistical models for other tasks, and we put the rest of this site together
in just a day---we really have nothing to lose.
I don't believe you!
Good! You should be skeptical when you approach any Red Balloon team!
So here's our best shot at convincing you: we are making everything 100% transparent. As
soon as you click on the map spotting a balloon, everybody on Twitter will know it.
Our only advantage
towards helping you win the prize is that we know which combinations have already been tried,
so we can dole out combinations systematically.
Ok, I'm sold.
Great. Get a guess to submit.
I'm still not sold.
Ok, well we tried, but there's still nothing to lose. You might as well get a guess to submit.
PS. How many guesses would it take to try all possible square mile balloon location combinations?
By our back of the envelope calculations, almost a googol.
Literally.
*Actually, if you do want to give us a cut, we're happy to take it. But you have
no obligation to give us anything.
