What streamers want to stream
See which game qualities streamers save, skip, and struggle to find.
September 2026 edition · First-party data through .
How to read this
A save signals real intent. In our backtest, streamers went live with saved games at 26 times the rate of skipped games, usually within days. How we know.
- Built from aggregate first-party saves and skips over the trailing 365 days, weighted by how deliberately streamers decide.
- Aggregate only. No individual streamer is ever named or identifiable.
- This is a demand read, not a forecast. It does not predict viewers, streams, or sales.
Methodology and definitions
- Calls are made against the catalog baseline and only flagged when they survive a multiple-comparison correction; everything else reads as typical.
- Qualities marked directional lean on a few big games or are not yet confirmed by the most recent weeks. Lead on the ones that hold up.
- An avoid share is the percentage of the declared-preference panel with an active matching avoid tag.
- Reach is the percentage of that panel remaining after active style and avoid filters are applied. A matching avoid tag excludes a game before recommendations are ranked.
- Save demand is the consideration-weighted response to first-party saves and skips inside StreamGist's preference-filtered serving environment. For an avoid-tagged quality, it is conditional on streamers who tolerate that tag.
Which qualities pull saves
How often streamers save a game carrying each design quality, weighted by how deliberately they decide, against the catalog baseline of 20.5%. Whiskers are 95% confidence ranges.
† directional: leans on a few big games or is not yet confirmed by the most recent weeks. Not enough decisions to call: Gambling mechanics.
View the gradient as a table
| Design quality | Weighted save rate | 95% range | Read |
|---|---|---|---|
| Chat-interactive | 28.1% | 26.4% to 29.9% | above |
| Horror | 25.6% | 23.2% to 28.1% | above, directional |
| Edgy / mature community | 25.3% | 23.5% to 27.1% | above, directional |
| Competitive | 22.8% | 20.6% to 25% | about typical |
| PvP-centric | 22% | 20.1% to 23.9% | about typical |
| Graphic / visceral | 21.4% | 19.9% to 22.9% | about typical |
| High energy / pace | 21.3% | 20% to 22.5% | about typical |
| Grind-heavy | 20.7% | 19.4% to 21.9% | about typical |
| Story-rich | 19.9% | 18.7% to 21.1% | about typical |
| Deep / study-heavy | 19.8% | 18.8% to 20.7% | about typical |
| Aggressive monetization | 18.6% | 14.7% to 22.5% | about typical |
| MMO-scale | 16.4% | 14.2% to 18.6% | below, directional |
| Gambling mechanics | n/a | n/a | not enough decisions to call |
The editorial side of the same pattern: why horror and edge work when risk has a frame, why the grind stalls, and why complexity needs an explainer.
What streamers weigh and decline
Some skips are quick scrolls. Others come from a streamer who stopped, weighed the game, and chose not to play it. Across the catalog, 14.9% of weighed decisions end in that deliberate pass. Qualities in the grind, complexity, monetization, and MMO structure cluster draw it more often (16.3%) than intensity, edge, and interactivity qualities (10%). A deliberate pass from a streamer who looked and chose not to play is a positioning signal Twitch never exposes; it only shows what already aired.
Ruled out up front
The share of streamers who filter each tag out of their recommendations entirely.
- toxic47% rule it out
- gambling37% rule it out
- pay to play36% rule it out
- pvp23% rule it out
- horror15% rule it out
- mmo14% rule it out
- violent14% rule it out
- grindy13% rule it out
Read the two lists together
Games with an edgy or mature community save above baseline even though toxicity is the most-avoided tag. Streamers who rule it out never see those games, so the save rate is measured among streamers who allow them. Demand for edge is real, it is just concentrated. The same mechanic applies to horror.
What streamers want next to what the catalog serves
Declared preferences against the number of profiled games carrying each style. The widest gap right now: 72% of streamers want chill games and the catalog carries 203.
- chill72% want it · 203 games
- energetic45% want it · 410 games
- interactive45% want it · 156 games
- competitive41% want it · 271 games
- immersive29% want it · 361 games
- educational10% want it · 494 games
Streamer demand, answered
What kinds of games do streamers want to stream right now?
In the September 2026 edition, games with chat-interactive, horror, edgy / mature community qualities pull saves above the catalog baseline of 20.5%, while mmo-scale games sit below it. The report refreshes monthly from first-party save and skip decisions.
How is streamer demand measured here?
From aggregate first-party saves and skips over the trailing 365 days, weighted by how deliberately streamers decide. Calls are made against the catalog baseline and only flagged when they survive a multiple-comparison correction. No individual streamer is ever named or identifiable.
Does high streamer demand predict viewers or sales?
No. A save predicts the saver's own streaming (in our backtest, streamers went live with saved games at 26 times the rate of skipped games), but this report does not predict viewers, streams, or sales. It reads intent, in aggregate, before anyone goes live.
Run the same engine on one game: yours
Want this view for one game? The Streamer Demand Read shows its reach, save rate against comparable games, and the qualities that help or hurt it.