Before delving into the topic proper, it needs to be understood beforehand that the precise ratio of fake reviews is uncertain. This is attributed to skepticism cast towards the tool Sheila Allen utilized to derive the original figures. What is not in contention is the fact a significant portion of them are indeed fake.
Initially, following an analysis of 1000 positive user reviews for The Last of Us 2 on Metacritic, Sheila Alien determined that 47% of them were fakes. This conclusion was reached through the analysis of their composition, which revealed 47% of the thousand analyzed were the same review posted under different accounts.
Since then, miss Allen has had her analysis critiqued by several individuals. Of importance, Youtube Ryan Kinel claimed to have double-checked her work and reached a different conclusion. What precisely was said is lost as the original tweet was deleted, but the exact figure was called into question.
Regarding the TLOU2 fake review data, some have made vids already based off it. Sorry. Props to Ryan Kinel for taking a moment to double check the data before making a video about it just because it suits a personal bias (and for giving me a heads up about the faulty data)…
— Sheila Allen 👽 (@Sheilaaliens) July 10, 2020
How did this happen? Simply put, sample sizes. Her original findings are correct. Of the thousand she crawled, 47% were indeed fake reviews. The problem is there are as of this writing 28,870 positive reviews. Depending on what order the reviews are presented in, any given day sampling of 1,000 will produce differing results.
This isn’t anyone’s attempt at being misleading; it’s just how statistics work. In fact, controlling the sampling size and location can allow a person to present whatever results they want to be shown. For example, political polls sample college indoctrinated individuals to show the US President gradually declining, despite his job performance metric remaining stable.
— Sheila Allen 👽 (@Sheilaaliens) July 5, 2020
What is known is a significant portion of the positive reviews the game received are fraudulent. Perhaps it isn’t 47%, but we’re looking at a double-digit percentage. What makes this significant is without these fake positive reviews, the user score would be considerably lower.
Though it is speculative, previously I have written how the HBO show deal was contingent on the success of the game. According to the director, there isn’t even a script for the show, and no production beyond the concepts green lighting was scheduled to take place until after the game’s launch.
It is in Sony’s best interest to inflate the user score. This is not surprising following the DMCA abuses the company undertook to squelch all discussion of the game’s leaks before launch. With units remaining on the shelves. Sony’s decision to tank their reputation over The Last of Us 2 will likely come to haunt them going into their next-generation launch.