The five-star review system has long been a staple of digital commerce, from app stores to e-commerce platforms. But as a recent case involving the RSS reader Current demonstrates, the system is fundamentally flawed. When Terry Godier launched Current, he encountered an unavoidable problem: anything below five stars is perceived as a disaster, leaving developers and users trapped in a cycle of inflated expectations and distorted feedback.

The issue is starkly visible in app store reviews. Many apps with a 4.0 rating or higher are actually harmed by 4-star reviews, even when those reviews are overwhelmingly positive in text. Users write comments like “This is my favorite app!” or “Gamechanger!” but assign only four stars, dragging down the average. This paradox highlights a mismatch between the numeric scale and user intent. The five-star system, originally designed to offer nuanced feedback, has become a binary: anything less than perfect is a failure.

The History of Star Ratings

Star ratings emerged in the early days of the internet as a simple way to aggregate user opinions. eBay popularized the system in the late 1990s, followed by Amazon and later Apple’s App Store and Google Play. The premise was straightforward: allow users to rate products or services on a scale of one to five stars, providing a quick visual summary of quality. Over time, however, the system evolved into a game of expectations. A 4.5-star average became the baseline for “good,” while anything below 4.0 was often dismissed. This inflation was driven by the psychology of rating: people tend to rate extremes, ignoring the middle grades.

In the early 2000s, Netflix used a five-star system for movie recommendations, relying on user ratings to power its algorithm. But Netflix later shifted to a binary thumbs-up/thumbs-down model, citing the limitations of star ratings. The company noted that users often rated movies inconsistently—a three-star movie might be considered good by one person but mediocre by another. This subjectivity undermined the system’s reliability.

The Current Dilemma

Current, a new RSS reader launched by Terry Godier, exemplifies the brokenness. Godier observed that many users gave the app four stars even while praising it effusively. In a public post, he noted that these four-star reviews are “actively harming the average” of the app, despite the intent being clearly positive. The app’s overall rating might hover around 4.2, but to many potential users, anything below 4.5 is suspect. This creates a perverse incentive for developers to beg for five-star ratings, as a single four-star review can undo the work of several five-star ones.

The problem is not unique to Current. Countless apps suffer from the same phenomenon. A 2023 study by the University of Michigan found that the average app store rating has been steadily climbing over the past decade, but the correlation between rating and actual quality has weakened. The study attributed this to “rating inflation” driven by user reluctance to give low scores and developer pressure to maintain high averages.

Psychological Factors at Play

The human brain is wired for extremes when it comes to rating scales. Research in behavioral economics, such as the work of Daniel Kahneman, shows that people are more likely to rate at the ends of a scale—either 1 or 5—rather than the middle. This polarization is partly due to the “peak-end rule,” where memory of an experience is dominated by its most intense moment and its end. For apps, a minor bug at the wrong time can trigger a 1-star rating, while a particularly helpful feature can lead to a 5-star rating. The 2, 3, and 4-star options become rarely used, making the system effectively binary.

Additionally, the social pressure to conform plays a role. Users worried about appearing too critical may inflate ratings, while others, feeling that anything less than perfection is a failure, penalize apps for minor flaws. This creates a feedback loop where the rating scale itself loses meaning.

Real-World Consequences

The consequences of a broken review system are tangible. For developers, a low average rating can kill an app’s discoverability on app stores, leading to fewer downloads and lower revenue. For consumers, inflated ratings make it harder to distinguish truly excellent products from mediocre ones that merely have a high rating due to luck or manipulation. In extreme cases, developers resort to purchasing fake reviews or offering incentives for five-star ratings, further corrupting the system.

The problem extends beyond apps. On Amazon, products with thousands of five-star reviews may still be low quality, while a product with a robust 4.2-star average could be vastly superior. The platform has attempted to combat review fraud, but the underlying scale remains flawed. Similarly, on platforms like Yelp and TripAdvisor, the five-star system leads to a “celebrity effect” where popular businesses maintain high ratings despite service issues, while newcomers struggle.

Attempted Solutions and Alternatives

Recognizing the issues, many platforms have experimented with alternatives. Netflix’s move to a thumbs-up/thumbs-down system simplified feedback and improved recommendation accuracy. Google Play and Apple’s App Store have considered removing averages altogether or defaulting to showing “most helpful” reviews. Some companies have shifted to “recommendation” systems, where users simply indicate whether they would recommend a product to a friend, without a numeric scale.

For app stores, one proposed fix is to weight ratings based on reviewer reliability or to use a Bayesian average to account for low review counts. Another idea is to show the distribution of ratings instead of just the average, so users can see how many people gave 5 stars versus 1 star. But these adjustments might not solve the underlying psychological bias.

Another approach is to eliminate ratings entirely in favor of curated editorial reviews or community-driven feedback. For example, some RSS readers rely on small, passionate user communities to provide detailed written reviews rather than stars. Current, as a niche product, might benefit from this model, where the quality of feedback matters more than a numeric score.

Yet the industry is slow to change. App store leaders are reluctant to abandon a system that users have grown accustomed to, despite its flaws. The five-star system is deeply embedded in our digital culture, from Uber rides to restaurant visits. Changing it would require a coordinated effort across platforms and user education.

In the meantime, developers like Godier are left to navigate a broken landscape. Current is a well-designed RSS reader, but its rating may not reflect that. Users who see a 4.2 average might scroll past, missing out on a genuinely excellent tool. The five-star review system, once a useful shorthand, has become a source of noise and frustration.

The case of Current is just one exhibit in a growing dossier of evidence. From streaming services to e-commerce, the system’s flaws are multiplying. Until a better alternative emerges, both developers and consumers must remain skeptical of the numbers that dominate our screens.

Source: The Verge News