MEMBER BLOG

Game Revolution Exposed: Haiku Fridays biased in favor of veterans!
Posted on Friday, July 29 2011 @ 18:57:36 Eastern

Fuck Haiku Friday

A veteran-biased scam

I’ll prove it with math
 
Haiku Friday was birthed from the moist mind-womb of Duke Ferris on April 17, 2009.

Having acquired beta invites for new MMO Free Realms, Louis C.K.’s emaciated twin brother elected to distribute them amongst the hoi polloi by ranking their worth as people.

He chose with the same metric God uses to decide who goes to heaven: the American haiku.
 
Four centuries after the Portuguese introduced firearms to the daimyo in 1543, the Japanese passed an even more destructive force on to Western culture. The American haiku, or “Baka na Amerikajin” in Japanese, consists of seventeen syllables spread across three lines in a five-seven-five distribution. The use of a season word is also considered standard practice, though masters will substitute an 'iron haiku ingredient' in its place.
 
But alas, though the haiku itself is the perfect measure of pureness in character, Game Revolution eschews judging on merit and instead unfairly favors senior members, specifically those from 2005-2007, when assigning prize recipients. Discounting the contests in which every entrant won by default or the victor was not publicly announced, one hundred sixty prizes have been awarded to eighty-four unique winners over one hundred contests. Consider those winners. First is LawnGnome, with wins in eleven individual Haiku Fridays, a confessed felon with a record as long as Arizona is filled with scorpions. His join date: April 25, 2007. Ranking second with seven wins is De-Ting, a forumer who famously convinced Blake Morse that surgically exposing and then swallowing his own intestinal tract would perpetually power his body. He joined Game Revolution on November 11, 2006. In third place with six wins is the abstruse Rinnon. Some sources claim he was the original one hundred fifty first Pokemon, while others describe him as “the idea of superluminal velocity made manifest.” Join date: November eighth, two thousand FIVE. Between them—13% of all discrete Haiku Friday prizes.

But why stop there? Let’s use a discrete probability density function--I mean, a bar graph--to view the percentage of total discrete prize winners from each year (the blue lines) as compared with the percentage of total unique entrances in each contest from each year (the red lines). The higher the blue line is from the red one in the years 2005, 2006 and 2007, the heavier those elitist *******s heading Game Revolution are biased against newer members.



Oh, that is just ****ing swell. Oh seven has more winners, oh six has more participants, and oh five is dead even! Between the years: a one-percent difference. But you sycophants are dumber than a lobotomized rock if you think I’m done yet. If I can show that the number of participants in each year is statistically different than the number of winners, then we know that, one way or another, Game Revolution has join date bias. We’re busting out the big guns—it’s statistics time.

We can examine the data with something called a ‘two-sample Chi Square test’. Put simply: a chi square test looks at the squared difference between the number of winners and participants in each year--weighted by sample size--and then adds all of those differences up. This number is the chi-square value. The higher the number, the higher the differences, and the more confident we can be that the distributions of winners and participants are NOT equal with respect to their join dates. In fact, statisticians can compare the chi-square value to a special chi-square distribution and state to the percentage point their confidence that two sets of data did NOT come from the same distribution. In non-******* speak, that translates to how certain we are that Game Revolution has been ****ing one group or the other of members over. In statistics, 95% is the generally accepted confidence level at which a statement can be credibly upheld. A 95% confidence interval for our specific data translates into a chi-square score of roughly 18.3. If the chi-square value we calculate for our data sets is 18.3 OR GREATER, then we will have statistically proven that the haiku crew at Game Revolution is lower than the boot scum on a shark-swallowed seafarer. Let’s do some math:



ARE YOU FUCKING KIDDING ME!? A FUCKING SEVEN-POINT-TWO!? Phoenix-****ing-Wright couldn't prove **** with a seven-point-two! Do you know how confident of Game Revolution's foul play we can be with a seven-point-two? 30% confident--not even a fifty-fifty shot. After all that work, there is literally LESS than no credible statistical evidence that Game Revolution has been doing anything but judging Haiku Fridays without considering join dates in any way.

 This is not to say that Game Revolution judges Haiku Fridays solely on haiku quality.

However, the weights and metrics that are used apply equally to every participant. At the end of the day, persistence and talent make consistent winners, not some arbitrary game state.

Don't believe it? Take a glance at this comparison of all time haiku winners and participants.

And from this, one truth:

You don’t win if you don’t play.

This was hard to write.

EDIT: A miscount! Chicanery! Outright bamboozlement! Taking a break from the exhaustingly rigid format imposed on the rest of the article, there was a mix-up in the Haiku Friday winners tabulation that assigned one of Lien's wins to De-Ting. This means that Lien is actually in second place with seven wins, and De-Ting is part of the triple-six trio.

Given that Lien's join date is 2008, catching this error before the launch of the article would have required major rewrites. How fortunate that I caught it just after... and had this revised winner's chart prepared in advance!