The Rating Rollercoaster

Developer Tools comes in at 0.9, roughly double the next tier down, where Business and Navigation sit at a steady 0.6. Beneath that is a long cluster of categories all parked at 0.5: Education, Sports, Utilities, Productivity, Stickers, Health & Fitness, Lifestyle, Food & Drink, Graphics & Design. Once you look at what these apps actually do day to day, the pattern starts to make sense. A developer tool can break the instant Apple or Android pushes out a new SDK, API, or OS beta. A patch usually lands within days, so the rating dips and climbs back within a single release cycle. Navigation and Business apps face a similar kind of exposure: a mapping API change or a payment-processor outage can tank reviews almost overnight, and a fix brings them right back up just as quickly. Nearly every category on this list shares the same underlying trait, a dependence on something the developer doesn't fully control, whether that's an external API, a seasonal spike in usage (Food & Drink, Sports), or a small review base where just a handful of ratings can swing the average (Stickers, Graphics & Design). If you're tracking app quality over time, this is worth keeping in mind: a single snapshot rating in these categories tells you almost nothing about where things stand next month.

Games comes in at 0.2, making it the steadiest category in the whole measurement, with News and Finance trailing just behind at 0.3 each. A wider cluster sits at 0.4: Shopping, Medical, Travel, Entertainment, Music, Reference, Photo & Video, Book, Weather. None of that has much to do with craftsmanship. It's really just volume. Games rack up huge numbers of reviews, and once an average is sitting on top of that many ratings, a handful of new one-star or five-star reviews barely nudge it. News and Finance apps behave the same way: established publishers, large user bases, and a steady trickle of reviews instead of sudden spikes. Compare that to Developer Tools at 0.9, and the gap looks like a difference in review-base size and review flow, not app quality. So if you're staring at a single rating for a Game or a Finance app, trust it. There's too much weight behind that number for it to move easily.

Paid apps swing 0.4. Free apps swing 0.4. Same number, right down to the decimal. So much for the theory that paid apps hold steadier because they attract a more committed reviewer, or that free apps bounce around more since anyone can grab one on a whim and leave a rating without spending a cent. Neither story holds up once you look at the numbers. What actually drives volatility is category and review base size, not whether money changed hands at install. If you're trying to guess how much an app's rating might move, forget the price tag. Look at what kind of app it is and how long it's been around instead.

Nothing else in this dataset moves like this gradient does. Apps released in 2026 swing by 1.4, apps released in 2025 swing by 0.5, and from there the numbers step down in order: 2024 sits at 0.3, 2023 and 2022 both land at 0.2, and everything from 2021 back through 2015 settles at 0.1. Really, it's just math. A brand new app might only have five or ten reviews total, so a single furious one star rating, or one glowing five star rating, can yank the average around on its own. An app that's been live since 2015, though, has years of reviews piled up, and no new review, or even a cluster of them, is going to budge that average much. So the 1.4 figure for 2026 doesn't mean newer apps are worse. It means their ratings simply haven't settled yet. Treat a rating on an app released this year as provisional, and treat a rating on a 2015-era app as about as settled as you'll find anywhere in this dataset.

Ten apps are deadlocked at the top of this study for the single biggest swing we found, a jump of 5, and together they account for 10% of the 50 apps in this wildest-swingers cut. The lineup is a bit of a grab bag: Gloaming: Generative Sunsets, Sbu Subscriptions, DashNote - Quick Notes, AnyTally: Counter & Tracker, Majella: NFP & Period Tracker, Ace AP STEM, Cratelog: Book Box Tracker, ShikiRip, Video Compressor: Reduce sizes, and Coffee Recipes: DupeBrew. No shared category. No shared price model. No shared audience. What ties them together is a thin review count, and that's really all it takes. Give an app just a handful of ratings, and one rough patch of reviews, or even a single bug fix update, is enough to flip its average from one extreme to the other. Once a review base grows to real scale, that kind of flip just can't happen anymore. So next time you judge an app by its star rating, keep this in mind: these swings aren't pointing to some hidden pattern. They're just small samples doing what small samples always do.

Look across 203,361 tracked apps and the average swing comes out to 0.42, which is just 8.3% of the maximum possible swing of 5. On its own, that number looks almost boring. But 20,900 of those apps, 10.3% of the whole set, land in what this study flags as a big swing. So one app in ten is dealing with genuine rating instability while the other nine basically sit still. The gap between that 0.42 average and the 5-point ceiling tells its own story: the distribution is lopsided, with most apps barely moving and a smaller, meaningful chunk swinging hard enough to pull the top end way out from the middle. So if you're judging app quality purely by average ratings, keep this in the back of your mind: the typical app is steady, sure, but that steadiness isn't universal, and right now something like a tenth of the market is going through real rating volatility.