Steam Refunds Controversy: Short Playthroughs Burn Indie Devs

There is a special kind of rage that comes from watching someone eat an entire meal and then send it back for a refund, claiming they didn’t like the first bite. That is roughly what indie developer Matteo “Zoroarts” Covic is accusing Steam players of doing to his game, Paddle Paddle Paddle-a cooperative boat adventure designed for roughly four hours of play, but speedrunnable in under two. On paper, his numbers look like a robbery: more than 55,000 refunds, a 21% refund rate, and “Very Positive” reviews all on a title that costs less than a latte during a Steam sale. But here is the uncomfortable part I keep coming back to: I have also been the person who bought a game, played ninety minutes, realized it was nothing like the trailer, and sent it back without a shred of guilt. Both of those people exist. The problem is that Steam’s refund policy pretends they are the same.

The $3 Dilemma: When Finishing the Game Fits Inside the Refund Window

Covic’s argument is brutally simple. Paddle Paddle Paddle is built around a tight, contained experience-about three and a half hours for the full level run, plus a demo that adds another forty minutes or so. At a $3 price point, there is barely any margin to absorb shock. When speedrunners or even just fast co-op pairs blast through the whole thing in one to two hours, they land squarely inside Valve’s refund eligibility: under two hours of playtime, within fourteen days of purchase. Covic claims this is being exploited as a free rental service, and the review sections he points to-where players mention knocking the game out quickly before refunding—certainly read like receipts.

Let’s be real about what that does to a small team. A 21% refund rate on a game that cheap is not an accounting footnote; it is an amputation. Indies live and die by conversion rates, and when a chunk of your “sales” are effectively zero-dollar loans from players who got the full experience, the math stops being math and starts being an exit sign.

But Steam’s Refund Policy Is Also the Only Reason I Trust the Store

Here is where I slam the brakes. Steam’s two-hour window is the single most effective consumer protection in PC gaming. It is the reason I will take a chance on a janky-looking genre experiment from a studio I have never heard of. It is the reason I do not feel like a mark every time I roll the dice on a co-op indie that might not even launch on my rig. Without it, Steam becomes a flea market where every purchase is final and every trailer is a pinky promise.

And not every refund is a heist. Some of those 55,000 are players who bought Paddle Paddle Paddle thinking it was a solo chill session, found out it demands coordination with a second paddler, and noped out. Some bought it for their kid, realized the humor did not land, and returned it. Some probably had controllers drift or frame-rate issues on older hardware. The policy does not distinguish between “I finished it and I’m scamming you” and “this is not what I paid for,” because from Steam’s perspective, the playtime is the only truth that matters. I do not blame Valve for that. Reading intent is impossible at scale. But that is also why the current system is a blunt object smashing both flyswatters and fine china.

Free Rentals vs. Buyer’s Remorse: The Line Steam Can’t See

The core tension is not whether refunds are good. It is whether a player who extracts the complete narrative and mechanical value of a sub-two-hour game before clicking “refund” is engaging in abuse, or simply using a system as designed. I think there is a difference, even if it is blurry. When I refund a game after an hour because the controls are janky, I am rejecting the product. When someone blasts through an entire curated four-hour experience in ninety minutes and then asks for their three dollars back, they are rejecting the price tag after consuming the product. One is a quality judgment; the other is a completionist speedrun of the honor system.

But how do you codify that without turning Steam into a surveillance state? You cannot just ban refunds for games under two hours, because that would make every sub-two-hour title a trap. You cannot ask developers to prove abuse, because they do not have Valve’s data. And you definitely cannot scale a policy that treats a completed short experience the same as a brief taste of something ten times its size. The homogeneity is the bug.

What a Fix Would Actually Look Like

If Valve wanted to seriously solve this instead of just weathering the PR cycle, there are smarter levers than scrapping the two-hour rule entirely. The most obvious is expected-play labeling. Steam already knows the median completion time for most games thanks to player data. If a title’s average finish is ninety minutes, Valve could flag that transparently on the store page—“Median completion: 1.5 hours”—and adjust refund messaging accordingly. Not to deny refunds, but to make it clear that finishing the game and returning it is visible behavior.

Another option is proof-of-play telemetry. Not invasive spyware, but simple checkpoints: did the player reach the credits? Did they unlock the final level? Steam already tracks achievements; tying refund eligibility to completion milestones for games with verified short runtimes is not technically radical. If you have seen the ending, you have received the core good. That does not have to mean zero refunds—technical issues and genuine disputes should always be winnable through support tickets—but it adds friction to the “free rental” pipeline.

Then there are repeat refunders. Steam already knows who chronically buys and returns games. A soft cap—say, flagged review after a certain percentage of refunded purchases—would not hurt honest buyers who get unlucky twice a year, but it would put a collar on the handful of users treating the store like a public library. The key is that any fix has to be surgical. Blunt policy changes punish the curious and the cautious, the exact people who make the indie ecosystem viable.

The Numbers We Have vs. the Numbers We Need

I would be remiss if I did not mention the grain of salt. The 55,000-refund figure and the 21% rate come from Covic himself. They are not independently verified, and “Very Positive” reviews mixed with massive refunds could just as easily indicate a game that hooked players emotionally but had technical barriers—co-op connection issues, controller mapping problems, or simple genre mismatch. I am not calling him a liar. I am saying that building a policy revolution off a single developer’s viral thread is how you end up with rules that hurt everyone else. If the data holds up across multiple short indies, then we have a trend. Until then, we have a warning shot.