Why Game Review Scores Don't Tell You What You Need to Know

A game sitting at 85 out of 100 across major review outlets tells you almost nothing about whether you specifically will enjoy it. That number is an average of opinions from people whose taste, patience, and preferred genres may have nothing in common with yours, compressed into a single figure that looks precise but hides more than it reveals.

Scores average away the information that actually matters

A game can score well because it nails atmosphere and story while having combat that a shooter fan would call mediocre. Averaged into one number, that nuance disappears entirely. A reader who cares mainly about combat and a reader who cares mainly about atmosphere would walk away from that same score with completely different expectations, and only one of them would end up satisfied.

This is why reading the actual review, not just the score, matters more for games than for almost any other kind of media. A 7 out of 10 from a critic who dislikes the genre you love can be a stronger signal than a 9 out of 10 from someone who loves everything in that genre indiscriminately.

Review timing changes what gets evaluated

Most reviews are written against a short review window, often under a week, sometimes played before day-one patches have landed. Games with slow starts, systems that only click after ten or fifteen hours, or multiplayer modes that need a populated server to actually shine, are structurally disadvantaged by this timeline regardless of how good the full experience eventually becomes. A score published in the first week is evaluating a different game than the one players experience a month later, once patches have landed and the community has settled in.

This matters most for anything built around systems that take time to reveal themselves, strategy games with deep mechanics, management sims, anything where the first few hours are deliberately slower than what comes after. A critic working against a deadline simply has less room to let that kind of game unfold at its own pace.

User scores solve one problem and create another

Aggregated user scores are sometimes treated as the more honest alternative to critic scores, since they represent a much larger sample of actual players rather than a small group of professional reviewers. That is true, but user scores come with their own distortion: review bombing around unrelated controversies, day-one technical issues that get fixed within weeks but permanently drag down an average, and a self-selecting audience that skews toward people with strong opinions in either direction rather than the quieter middle who simply enjoyed the game without feeling compelled to rate it.

Neither critic scores nor user scores are lying exactly, they are each measuring something real, just not necessarily the thing you actually want to know before spending money and time on a game.

What actually predicts whether you will enjoy something

The single most useful piece of information is rarely the score itself, it is whether a reviewer's stated preferences overlap with your own. A reviewer who explicitly says they bounced off slow-burn RPGs and then gives one a mediocre score is giving you useful information if you love slow-burn RPGs, it just means read that score as a point in favor, not against. Finding two or three critics whose taste has historically lined up with yours and following their specific reviews, rather than an aggregated average from dozens of strangers, produces far better predictions than any single number.

This same logic applies to how to read a PC build guide without overspending, the specific reasoning behind a recommendation matters more than the headline number, whether that number is a benchmark score or a review score.

Where I disagree with how aggregation sites present themselves

Review aggregators present themselves as a neutral, objective summary of quality, which flattens a genuinely subjective medium into something that looks scientific. A restaurant critic's 8 out of 10 and a combat-focused player's reaction to the same game's combat system are not measuring remotely similar things, yet aggregation sites display them side by side as if they are comparable data points feeding into one true number.

I would rather see review aggregation disappear entirely in favor of reading two or three trusted voices directly, but since that is unlikely to happen given how convenient a single number is for marketing and purchasing decisions, the realistic fix is treating the aggregate score as a rough filter rather than a verdict. Anything above a very low threshold deserves an actual look at what specific reviewers said, not a purchase decision based on the number alone.

A better way to use scores without being misled by them

Use the aggregate score to filter out genuinely broken or poorly received games, then switch entirely to reading two or three reviews from critics whose taste you have learned to trust over time. Pay closer attention to what a reviewer specifically praised or criticized than to the number they attached to it. A detailed complaint about pacing in the first three hours matters more to your decision than whether that complaint cost the game a seven instead of an eight, and that distinction is exactly what a single aggregated score can never tell you.