Share this page with other Digg Readers or tell others you "Dugg" it.
Share on Reddit
Share this page with other Reddit Readers or delegate points for this page.
Woh! What is this thing?!
This is a social bar from the Dilbert team and our friends at Meebo. It is a cool little feature we've added to Dilbert to make sharing comic strips, Mashups and other content, a simple click of the mouse from any page.
Try it out – use the buttons to instantly send the next funny comic you read to your Facebook status or your Twitter feed. We’re adding even more ways to share soon, so check this space often!
@MarillionFan and ihateregistering: It's a comic. I found it funny. I read your comments. The following happened: confusion setting in...processing logic...trying to find relevance in comment...processing....processing....processing...head about to explode....nine nine nine nine nine nine nine nine...
@cerebellum, in addition to what has already been said, you are also wrong by a factor of ten.
The first digit could have been anything. You only see it odd that the second to sixth digits are the same, and that occurs once in 1 in 100,000 times.
Taking ihateregistering's suggestion of a digit a second, the RNG will, on average, give a number six times in a row once every 27.7 hours, so about once a day.
@cerebellum, the problem with that reasoning is that the probability of ANY six-digit sequence is 1/0.1^6. But if the RNG had tossed out 340923, you probably wouldn't think it was biased, even though the probability of it coming up is exactly the same as 999999 in a random system. 999999 is only suspicious because we see a pattern in it assign it special meaning, so the real test is to consider what percentage of the numbers from 000000-999999 we've assigned special meaning. What are the chances of getting a number where we see a pattern vs one that "looks random"? If the RNG is really that broken there's no reason to assume that each digit is generated separately, so it's not just 000000...999999 that we have to look at. There are obvious patterns in 123456, 654321, 696969, 246810, etc. More observant or mathy types might be struck odd by numbers like 314159, 112358, 357911. Or numbers that have special meaning in the current context, like what if 459-834 is the building and room number of the accounting office, wouldn't a generated number of 459834 seem a little unlikely? What if it spat out 121501 at 12:15:01, or 042010 on April 2010?
Now, if BEFORE you started running the RNG, you were looking specifically for the sequence 999999, that would be different. But finding a pattern in numbers after the fact isn't special; our brains are wired to see patterns even when they aren't there.
Dilogic's explanation of why the comic is funny is correct. The RNG seems quite obviously broken, and it probably is, but without cracking open his head and looking at the source code, there's really no way to know for sure.
