Steam's Sloppification Investigation

1. Release volume over time

Dated Steam game releases only. Per-day uses 365.25.

Percent YoY is useful, but it lies by omission when the base keeps growing. 2014’s +232% looks like the monster of the series; recent years look calmer. So here is the same years again as absolute titles added - how many more games than the year before. That is why this follow-up exists: to put the relative spike next to the pile that actually hit the store.

The yearly count has been compounding hard for a decade - release volume roughly doubles every few years. If that keeps up, the late decade is not “a bit busier.” It is another step-change in daily volume. Last twelve months and 2026 year-to-date already sit on that hotter pace.

So you think it’s bad now? We ain’t seen nothing yet. This release rate is off the charts. And this is just on Steam. I imagine the volume situation on itch is already a lost battle - might poke around their data later.

2. Attention concentration

Steam has to make money. With volume getting worse, they need to tune discovery toward what is “safe,” what converts better. They can’t be showing “Jeffrey Eggstein” to everyone all the time, so they churn attention on risky projects faster and keep focus on what is selling. Which makes it feel like we only see the same games all the time - and when we see something new, maybe it’s earnest work fighting for a slot, maybe it’s slop being tested for engagement. Still to be seen either way.

Over time that concentrates attention on the few games that are already proven. Finding hidden gems is not something Valve can prioritize as a business; it would be too wasteful for them. Storefronts also have to protect sales and filter risk, so they cannot endlessly spotlight unknowns. Players feel that as spam, filler, and the same famous names crowding every shelf.

Among games that already have enough reviews for Steam to show an official score band - the visible mid-and-up tier, not every silent listing - the concentration is brutal:

A tiny slice of the scored set owns most of the review mass. The long tail barely shows up in that reception data at all. Most of the catalog is not “bad by default.” It is just hard to notice next to the pile that already has momentum.

Tens of thousands of titles still have little public signal - and a large chunk have no Steam review signal row in this database yet.

Popularity feeds itself. A game that gets shown gets reviews. Reviews make it easier to trust. Trust makes it easier to show. A studio with a launch push can buy the first shove into that loop. A solo developer often gets one noisy week and no second chance.

3. Limits of store tags

Tags are useful until they are not.

“Show me indie action” is still a haystack. Tags describe marketing buckets more than they describe how a game feels. That is fine for filters. It is a weak substitute for “find me something like this, but smaller and stranger.”

It is not enough to use predefined tags - or any tags at all - if you are a game developer. Relying on Steam’s SEO is a death sentence. The wishlist game everyone is playing is a good move, but you need to keep pushing for months after release and really focus on building communities outside of Steam to survive the great slop.

4. Implications for players

If you mostly use the store search and the front page, you are fishing in the brightly lit part of the pond. The catalog is not empty of good games. It is drowning them in volume, while a thin slice of titles absorbs most of the social proof and organic traffic.

That is why discovery feels broken even when the store is full.

The other half of the problem is how we search. Feeling-based questions (“cozy deckbuilder nobody talks about,” “games like this but smaller”) do not map cleanly onto tag checkboxes or chart position. Matching on what a game actually is - synthesis from listing content plus player-signal keywords, similarity shelves, vibe-style routing - is a different job than ranking what already converted. Rank on fit, and when relevance is close, do not let brand gravity and review volume win by default. Soften the mid-size / AAA shove on ranked shelves. Treat borrowed storefront fame as something that can demote spam, not buy a higher slot. That is the parallel lane I care about: not pretending storefronts are evil, just refusing to rebuild their incentives.