how true is this (from a reddit comment about a post on KSP steam rating page): That…
- User
how true is this (from a reddit comment about a post on KSP steam rating page):
That reminds me of the british gaming magazines of the 1990s. Basically you'd subtract 90 from the score and multiply the remainder by 10 to get a better feeling what the true review percentage was. A 91% or 92% was an indicator of a truly awful game. But they all advertised in the magazine so they couldn't say that out loud.
Assistant
This comment contains a kernel of truth about gaming journalism in the 1990s, though it's somewhat exaggerated.
During that era, gaming magazines did face real conflicts of interest - they relied heavily on advertising revenue from the same publishers whose games they reviewed. This created pressure to avoid harsh reviews that might jeopardize those advertising relationships. British magazines like PC Gamer UK, Edge, and others were part of this ecosystem.
The scoring inflation was real - many magazines clustered their reviews in the 70-95% range, with scores below 70% being relatively rare even for mediocre games. A game getting 91-92% often was indeed just "decent" rather than exceptional, and truly outstanding games would get 95%+.
However, the specific formula mentioned (subtract 90, multiply by 10) is more of a humorous exaggeration than a literal truth. A 91% game wasn't "truly awful" - it was usually just average or slightly above average. Genuinely bad games did sometimes receive scores in the 60s or below, though less frequently than they probably deserved.
The situation has parallels today with concerns about review embargoes, early access, and maintaining relationships with publishers, though the dynamics have evolved with the rise of independent media and user reviews. The core issue of financial incentives potentially influencing coverage remains a topic of discussion in games journalism.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.