American Idol - judges' votes

Does anyone keep track of all votes made by the judges in American Idol TV show? It seems it would be very interesting to apply some statistics to find out some patterns. This ocurred to me when in the present season, Paula Abdul voted "no" when both Simon Cowell and Randy Jackson voted "yes" for one candidate in Dallas. I couldn't remember whether this has ever happened.

I came up with the following hypotheses:

1. - Simon's "yes" vote is a good predictor of acception unanimity - if he votes "yes", everyone else will likely vote yes.
2. - Paula's "no" vote is a good predictor of rejection unanimity - ig she votes "no", everyone else will likely vote no.
3. - Randy's vote is very correlated with Paula's vote. They have a tendency to vote the same.
4. - The order in which judges vote affects the final outcome.

Of course, what is aired is only a subset of the whole show and surely biased by th editor to smooth any prejudice or repetitive pattern. I don't know how to take that into account. But nevertheless, is there anyone willing to watch all seasons again and account for every vote? :)

Elf Yourself

To make yourself look like an elf seems to be becoming a christmas tradition - every year now a few dozen friends "discover" this popular, yet cool, site where you can upload a photo and it will be transformed into a dancing elf.

The site uses flash to handle the dancing elf and is pretty neat. Interestingly, you may create elfs only during christmas time - in february the site closes only to probably reopen in december again. So, hurry.

From the series "Interesting Mindgames"

This is a nice game to play in a class in order to try to put students' brain into first gear. Ask everyone to write down on a paper a number between 0 and 100, but mention that they must try to write a number that is as close as possible to 80% of the average of the numbers chosen by the group.

This is likely to fall into a normal curve, skewed towards smaller numbers. If one was playing alone and should guess the average of random numbers, one would guess 40 (considering a minimum knowledge of probabilities). But 80% of 40 is 32 and another student could reason that the first student would choose 40 and, thus, he should write 32. The first student could then reason that the second student would reason like that and choose 80% of 32: 25.6. And so on.

If everyone would reason like this, they all would reach the conclusion that they should choose zero, because everyone would choose zero. This is the Nash equilibrium of the game. However, no group of student that have never played the game would have a majority (or even more than a few) choosing zero. When trying to anticipate other's choices, it seems humans consider other human's behavior to be completely random.

The interesting thing is that if someone does choose zero, he would probably lose the game, despite being "right". He is right from a game-theoretical point of view but plain wrong from a psychology point of view.

Windows 386 Video

This is pretty funny, a long promotional video of Windows 386 (the successor of Windows 2.0, but not quite Windows 3.0 yet). Even considering it was made in the late 80s, it is pretty bad. It was not a video that was shown on TV, since it is over 14 minutes long, but was intended to be distributed to retailers. Jump at least the first 5 minutes of the video, otherwise you may die of boredom.

Double Slit Nightmare

Things like this keep me from sleeping at night. The quantum double slit experiment is freaky. But it could be put to good use, if only we accept it as correct (as I never will).

For instance, we could test cryptography algorithms. We record the information of which slit the light/electron is passing through, but instead of erasing it (to create the fringes), we encode it using some cryptography algorithm. IF the algorithm is unbreakable (provided we have erased the random-generated password), then we should see fringes. However, the experiment would not mean that the algorithm is really unbreakable - it would only mean that no one will ever break it, either because no one will ever try or no one will be able to. Either way, data should be safe forever.

Having a password to decrypt the information should also be useful for other things. Say we create a password, but never look at it. We beam the password to space by using radio signals and then erase any local version. What should the experiment look like? If there are fringes, it means we'll never recover the password or the universe (or we) will be extinguished before we are able to break the code. But it there are no fringes, this means password and data will be reunited! This could be done by many ways - we could in the future be able to travel faster than light, the beam could bounce back or an alien race could pick up the signal and have access to the data eventually. But if there are fringes, this means that we'll never recover the password. Indeed, if there are fringes, we would never try because we would believe it's impossible! Our beliefs can change the experiment's result, not only our knowledge.

The quantum double slit experiment can be used to predict some possible futures, or at least rule out possibilities. Such futures will be attached to our beliefs and knowledge, but are futures nonetheless. And predicting something we create should be a good thing. Or not.

Making your own Lava Lamp

For some time now I've been wanting to create my own Lava Lamp. Actually, I tried only once. It seemed pretty simple by looking at a commercial lava lamp, so I gave it a try. I just got some paraffin wax and put into a glass of water. Then I heated the bottom and waited to see those lovely bubbles.

It couldn't be more simple. And it wasn't.

The paraffin wax melted, alright, but no bubbles were formed and I ended with a strange sculpture of solid wax that was neither beautiful nor artistic. Wish at that time I had a camera to show the mess.

Anyway, since then I found some nice sites on the web that try to teach how to make your home-made lava lamp. It's far from simple and reading those made me feel my commercial one was pretty cheap after all. Here follow some links:

• http://www.factmonster.com/spot/lavalamp.html
• http://www.oozinggoo.com/howto.html
• http://chemistry.about.com/cs/howtos/ht/lavalampreal.htm

In the Wikipedia there is an article on lava lamps that is instructive to understand why you're following all those instructions and nothing seems to work.

Name That Game Audio Challenge

One thing we may always count on is nostalgy. The quiz game "Name that Game Audio Challenge" appeals to anyone missing the old games of the 80's, by challenging the player to guess which game a particular sound belongs to.

I did okay in Atari games, but ironically newer games were harder for me. All in all I got 9 of 18 right answers and found myself wondering how I enjoyed so much those 4 colors 2 tones games...

Peekaboom

After the SETI Program requesting the power of our personal computers to search for extraterrestrial intelligence, it is now time for a project to request the power of our own brains to develop artificial intelligence.

This is the goal of Peekaboom. From the outside, it is a simple game of guessing the picture. Players play the game in pairs and one is given a picture that the other must guess. To do this, the first player chooses parts of the picture to be shown. Both players receive better rewards when fewer parts are shown.

While the game itself is quite enjoyable, it is its purpose that makes it stand out. When players are playing, an underlying algorithm is constantly observing what people consider useful information in a picture. This, in turn, will be used to design better artificial intelligence algorithms, which may use an approach similar to the one we use to understand images in the real world.

CiteULike

A very good idea: CiteULike. It is an online organiser of academic papers and citations. But rather than just keep record of your papers, it also works in a "social network" kind of way. You can see other papers that are related to yours and even most popular topics.

It makes easy to keep track of references, since many sites, including Amazon, offers a link to add an item directly to your list in CiteULike. Also, if someone else has registered some paper or book, you may just copy the reference to your list. No need to search for authors, dates and editors.

Bud Light's Real Man of Genius

A nice set of radio spots for Budweiser Light won the Cannes Lions award in its category. Really worth listening to each one. You may download those and much more here.

Made to go with the TV spot from Honda, also presented in Cannes, the Grr Game. Completely non-sense and very well done in Flash.

Gathering for Gardner Paper Dragon

One of the most powerful optical illusions I've ever seen is Jerry Andrus' paper dragon. It is a dragon made of paper which is folded in a certain way that it seems to follow you with his head wherever you go. Very interesting.

(note to notpron players: yes, it is the same dragon. perspective is the key.)

The project of this dragon is very easy to assemble. It is not an origami, all you have to do is cut on the template lines, fold it and glue some parts. It usually does not work with both eyes on short distance, you may get a better effect by either closing one eye or looking at it from some distance (2 meters is enough). Also, light influences the effect, it is not good to have a direct light spot over it - the less shadows the better. Check this video to see how it works.

You may use Photoshop to change the color of your dragon, mine is green (dragons are green, aren't they?). I have made templates available in three different colors:

Green Dragon, Blue Dragon, Red Dragon

There are many optical illusions on this site, where the original template came from, many of them are viewable directly through the browser and others require some assembling.

Update: a commenter found a different version of the same illusion, with a skeleton instead of a dragon. I didn't assembly it yet, but it seems to have far less folds than the dragon, which may compromise the effect but make it easier to make it. You may find it here.

10x10

TenByTen is something like a resume of the news on the internet at any given hour. When you access the site, you're given 100 words and 100 pictures in a 10x10 grid that are the most frequently cited among their sources.

The images get repeated frequently, but that's more or less wanted since it is a way to show that some image is being very cited and, thus, important for the time.

The interface is in Flash and new data is retrieved every hour by the server, so it is not done on the fly. The idea is a very clever one and well implemented.

Straight Line Design

Some of the funniest furniture in the world: Straight Line Design. There is no price in the site, but I bet they are expensive. Imagine a whole house full of those things.

Nice Flash site, also.

43things

43things is an interesting twist to on-line communities like Orkut. When creating a profile, you define 43 things you want to do, including desires, wishes or plans. Then you are connected to people wanting the same things. People may comment on the items in the list and group into communities to exchange ideas on how to accomplish things.

The idea of the author, an ex-employee from Amazon, is to allow people to know what things are being wanted throughout the world and help them accomplish those things. Even if it is to take over the world.

Peasant´s Quest

Still in the realm of adventure games on-line, there is the Peasant´s Quest, which is a Flash-based game that emulates those old Sierra style adventure games. The game is ugly, but is meant to be like this. There was a time where we only had 16 colors and 320x200 pixels to work with...

The game is not very long and is very funny. You may even save your progress, so its a good candidate to a game playable at work's breaks.