...to have one entry a month. Oh wait. I did that. (Actually, no. I was just testing backdated entries.)

'Nuff said on how the Internet changes lives. Err, or not. I'll say this once more: the Internet has changed our lives. Well, at least mine and those of the people I know.

Thanx to the Internet now everyone is an expert. Oh, so you thought you'd found some piece of hot information and perhaps you would share it by forwarding to all the mailing lists you knew -- and earning a few credits in the process? I bet within minutes some guys would a) declare it a hoax by virtue of several URLs attached; or b) support it by providing a much more comprehensive compilation -- stealing most of your credits.

You know what, I saw Jesus today at lunch. Yeah, even the Son of Man needs to eat real food once in a while. Man, that guy looked much like Him -- the hair, beard, skin color, and all -- that were there a Jesus-look-alike contest, he would be the winner in no time and Our Savior would come second. All right, a very close second.

Oh, and he was clutching -- get this -- a Cisco router! I know! How fitting, right? Uh, you didn't get that? Never mind then.

Anyway. Let's switch to the IT-freak mode. After getting myself frustrated over this Monty Hall thingy, I finally decided to come up with an experiment of my own. You know, just to be sure that those simulators elsewhere were not rigged or anything like that. (Which, of course, they were not.)

And just to be as sure that I didn't make stupid mistakes in writing my simulator, lemme tell you how it works. Hey, who knows, you might thank me later for making you fall asleep.

First things first, I use a manual seed for the random number generator on purpose. So that if anyone finds anything interesting, it'll be easier to replicate the figure using the seed.

Some legend: C shows where the car is; red cell shows the original door selected.

Call it Siwis' fav pseudo code if you will, here's how I do it:
1) Randomly determine the door behind which the car is: car_idx.
2) Randomly determine which door is originally selected: select_idx.
3) Randomly determine which door with a goat behind it is to remain closed: closed_idx.
4) "Open" the rest of the doors, leaving only car_idx and closed_idx closed -- one of these is of course select_idx.
5) If car_idx is the same as select_idx, then the car is behind the originally selected door and you would win if you stayed with that door; else you would win if you switched door.

The result? I still find it hard to believe it's not a 50:50 chance. I mean, to me it still sounds like I give you three cards face down -- two kings and an ace, ask you to pick one which you think is the ace, then throw away one of the kings, practically giving you a 1-in-2 shot. Right? Apparently not.

Current music: Bunglon - Denganmu
Current mood: happy