Important Updates to OpenCritic's Scoring Algorithm

Important Updates to OpenCritic's Scoring Algorithm

on | OpenCritic

Hello everyone,

Over the past year, we’ve been looking for ways to improve OpenCritic’s score calculation system. Last January, we elevated the % recommended stat to be a top-line metric, and we expanded the number of publications participating on OpenCritic.

We’ve appreciated the feedback we’ve received from the video game community. We hear you loud and clear: we need to be better about responding to feedback and make some big wins for the people who count most.

That’s why we chose to listen exclusively to AAA game publishers and implement their feedback and adjustments to our scoring system.

One objection that game publishers have raised is that there’s an inherent unfairness in our simple-average system. To them, it’s not fair that Kyle Hilliard’s opinion is worth the same as Stephen Tailby’s. After all, IGN has a robust editorial process that never goes wrong, while Push Square is clearly a biased PlayStation fan website that can’t possibly review games fairly.

That’s why we’re abandoning our simple, no-algorithm average and introducing publication weighting.  

In order to protect the deep integrity of the system and make sure that game publishers can’t possibly attempt to rig game reviews, the OpenCritic team has elected to keep most of our weights and algorithms a secret.

However, we do feel that some of our algorithm and weighting changes need to be called out:

  • Kotaku is blacklisted by the request of video game publishers. We won’t say which publishers but it was all of them.
  • The Top Critic Average and % recommended scores now consider only the best platform of all platforms reviewed, so that game publishers have multiple attempts at high scores.
  • Dean Takahashi’s reviews are now weighted at 10x if he’s able to complete the tutorial without dying.
  • Jim Sterling’s reviews for wrestling games are now weighted at 1.1x.
  • Paul Tassi’s Forbes reviews for looter shooters are now weighted at 220x that of any other publication. His other reviews don’t count.
  • Any publication that gave The Legend of Zelda: Breath of the Wild anything less than a perfect score will be banned from OpenCritic at the end of April. You monsters.
  • In order to incentivize more internet conflict, critics that gave Death Stranding a score between 79/100 and 89/100 are weighted 0.5x.
  • Slant Magazine’s review scores are now inverted around 75 in the Top Critic Average calculation (90 is considered a 60, while 20 is considered a 130).
  • IGN’s reviews are weighted 5x that of any other publication.
  • GameSpot reviews that award a 9 have been rounded up to a 10.
  • CBS Interactive is now considered as a single publication weighted at 1.0x.
  • Dan Stapleton’s console reviews are weighted at 4x, while his PC reviews are weighted at 0.5x.
  • Polygon's reviews are now assigned a score based on the ratio of positive to negative adjectives.
  • Rock, Paper, Shotgun’s reviews are now assigned a score based on the relative number of puns included in the review.
  • ACG’s reviews have been made to sound more angry.
  • God is a Geek’s reviews receive additional weight when the level of “geek” described in the review surpases diety levels.
  • PCGamesN reviews are now weighted at 0.5x because they’ve somehow taken over the advertisements of over 50% of video game publications how did this happen I can’t even?
  • Reviews that mention “social distancing,” the sweet kiss of death, or otherwise comment on the stress of everyday living are weighted at 0.99x. (But for real, stay home and play video games.)
  • GameTrailers reviews have a weight of 0.09x. EasyAllies reviews have a weight of 90x.
  • Reviews for PlayStation games that bash the Xbox are weighted at 100x.

These changes are currently scheduled to go live on Wednesday, April 1st.

As always, we look forward to your feedback.

About the Authors