Phone Detective

A study of the hidden formula that drives headlines Metacritic done this week, but

Kotaku

discovered some critical errors that challenge. earlier this week, Full Sail University Professor Adams Greenwood-Ericksen GDC held a meeting in San Francisco, where he shared some of his research on the effects of Metacritic, the site of aggregation Reviews takes hundreds of media outlets and as a single number, or Metascore. Metacritic has taken some heat in recent years for refusing to reveal the formula they use to produce their results. There is a simple way: Metacritic admits that give more weight to certain media during crunch the numbers. But I never said that weighting system works.

So when Greenwood-Ericksen said he had a model that Metacritic scores replicated, people realized. Gamasutra published an article entitled "Metacritic weighting system revealed," and we had a lot of game developers and journalists who speak. POS system classified into six different "levels" and gave great importance to sites like

IGN

and Wired (and significantly less weight than other major sites such as Giant Bomb

Shortly after, came out firing Metacritic. It took Facebook to bring down the formula, which he called "wild, totally inaccurate" and accused Gamasutra

running a misleading title. ( when

Kotaku for comments Gamasutra editor Kris Graft apologized: "Yes, I think that the main issue was a poor way, and we apologize for any confusion about this. was also unfortunate that a session with inaccurate because there was in the show. ")

Some, however, remained skeptical about the allegations of Metacritic, as the reader still does not share the formula they use.

But today

Kotaku

discovered a flaw in the formula of Greenwood-Ericksen: at least two weights for outputs

and

are incorrect.

start at the beginning. Greenwood-Ericksen model developed from data Metacritic from 2006 to 2005 or 2011-attributes certain numerical weights such as 1.5 and 0.5, each point of sale of video games. Formula: see page Metacritic video game, take all test scores listed are multiplied by the weight associated with each of its output, add them all, and divided by the total number of accounts. This model has been successfully replicated something like 50 partitions, Greenwood-Ericksen said.

Except that while plugging in the numbers and formula test today, I discovered that the math does not work for the PS3 game

Swords & Soldiers. When I tried to get the Metascore, I discovered that my results were 7-8 marks off. (The calculation worked for some of the other games that have experienced as Venetica