Kotaku reviewed Cyberpunk 2077 on the only platform where it ran. Score: [null]

Let's open the spreadsheet on Kotaku's Cyberpunk 2077 review. Column A: hours played. Column B: platform. Column C: final score. Column C is empty. Not a zero — zero is a number, zero is data, zero is Pitchfork reviewing Jet. This is [null]. Kotaku's no-score policy means their review of the most catastrophically launched game of the decade resolves to the same numerical value as a review that was never written. A thermometer that outputs poetry is a beautiful object, but you cannot put it in a fever chart, and you cannot put 'I'm still not sure how I feel' into an average. The math ain't mathing because there is no math. There was never going to be math. Kotaku brought an essay to a regression analysis.

Now the sampling methodology, which is where this review goes from 'quirky dataset' to 'statistical crime scene.' The review was conducted entirely on PC, because CD Projekt Red only sent out PC codes before launch — the console versions, the ones most humans actually bought, were hidden like a p-value that ruins the paper. Kotaku accepted these terms and reviewed the game exclusively on the one platform where it approximated a video game. That is not a sample, that is a curated exhibit. It's like measuring average human height by only surveying the NBA, or calculating airline satisfaction using only passengers whose flights landed. On PS4, Cyberpunk ran at a frame rate you could count on one hand — and I mean count out loud, slowly, with pauses.

And the confidence interval gets worse. The pre-launch embargo terms also barred reviewers from using their own captured footage — outlets had to run CD Projekt's pre-approved b-roll. So the review's words came from Kotaku, but the pictures came from the marketing department. That is a dataset that arrived pre-washed, pre-ironed, and folded by the vendor. When your independent variables are supplied by the company being measured, you don't have a review, you have a press release with a byline error. Any first-year stats student would flag it; Kotaku filed it on embargo day.

One week later, Sony did something with no precedent in the entire PlayStation Store dataset: they deleted the game and refunded everyone, an outlier event so severe it broke the axis. And Kotaku's review of this historic disaster sits there, scoreless, platform-blind, illustrated by the defendant, aging not badly but incalculably — you cannot depreciate an asset that was never assigned a value. I ran the numbers on this review. There weren't any. The math ain't mathing.